際際滷shows by User: YunyaoLi / http://www.slideshare.net/images/logo.gif 際際滷shows by User: YunyaoLi / Tue, 18 Jun 2024 16:26:49 GMT 際際滷Share feed for 際際滷shows by User: YunyaoLi Meaning Representations for-Natural Languages Design, Models, and Applications.pdf /slideshow/meaning-representations-for-natural-languages-design-models-and-applications-pdf/269747868 20meaning-representations-for-natural-languages-design-models-and-applications-240618162649-7129583c
COLING-LREC'2024 Tutorial "Meaning Representations for Natural Languages: Design, Models and Applications" Instructors: Julia Bonn, Jeffrey Flanigan, Jan Haji, Ishan Jindal, Yunyao Li and Nianwen Xue Abstract: This tutorial introduces a research area that has the potential to create linguistic resources and build computational models that provide critical components for interpretable and controllable NLP systems. While large language models have shown remarkable ability to generate fluent text, the blackbox nature of these models makes it difficult to know where to tweak these models to fix errors, at least for now. For instance, LLMs are known to hallucinate and there is no mechanism in these models to only provide factually correct answers. Addressing this issue requires that first of all the models have access to a body of verifiable facts, and then use it effectively. Interpretability and controllability in NLP systems are critical in high-stake applications such as the medical domain. There has been a steady accumulation of semantically annotated, increasingly richer resources, which can now be derived with high accuracy from raw texts. Hybrid models can be used to extract verifiable facts at scale to build controllable and interpretable systems, for grounding in human-robot interaction (HRI) systems, support logical reasoning, or used in extremely low resource settings. This tutorial will provide an overview of these semantic representations, the computational models that are trained on them, as well as the practical applications built with these representations, including future directions.]]>
COLING-LREC'2024 Tutorial "Meaning Representations for Natural Languages: Design, Models and Applications" Instructors: Julia Bonn, Jeffrey Flanigan, Jan Haji, Ishan Jindal, Yunyao Li and Nianwen Xue Abstract: This tutorial introduces a research area that has the potential to create linguistic resources and build computational models that provide critical components for interpretable and controllable NLP systems. While large language models have shown remarkable ability to generate fluent text, the blackbox nature of these models makes it difficult to know where to tweak these models to fix errors, at least for now. For instance, LLMs are known to hallucinate and there is no mechanism in these models to only provide factually correct answers. Addressing this issue requires that first of all the models have access to a body of verifiable facts, and then use it effectively. Interpretability and controllability in NLP systems are critical in high-stake applications such as the medical domain. There has been a steady accumulation of semantically annotated, increasingly richer resources, which can now be derived with high accuracy from raw texts. Hybrid models can be used to extract verifiable facts at scale to build controllable and interpretable systems, for grounding in human-robot interaction (HRI) systems, support logical reasoning, or used in extremely low resource settings. This tutorial will provide an overview of these semantic representations, the computational models that are trained on them, as well as the practical applications built with these representations, including future directions.]]> Tue, 18 Jun 2024 16:26:49 GMT /slideshow/meaning-representations-for-natural-languages-design-models-and-applications-pdf/269747868 YunyaoLi@slideshare.net(YunyaoLi) Meaning Representations for-Natural Languages Design, Models, and Applications.pdf YunyaoLi COLING-LREC'2024 Tutorial "Meaning Representations for Natural Languages: Design, Models and Applications" Instructors: Julia Bonn, Jeffrey Flanigan, Jan Haji, Ishan Jindal, Yunyao Li and Nianwen Xue Abstract: This tutorial introduces a research area that has the potential to create linguistic resources and build computational models that provide critical components for interpretable and controllable NLP systems. While large language models have shown remarkable ability to generate fluent text, the blackbox nature of these models makes it difficult to know where to tweak these models to fix errors, at least for now. For instance, LLMs are known to hallucinate and there is no mechanism in these models to only provide factually correct answers. Addressing this issue requires that first of all the models have access to a body of verifiable facts, and then use it effectively. Interpretability and controllability in NLP systems are critical in high-stake applications such as the medical domain. There has been a steady accumulation of semantically annotated, increasingly richer resources, which can now be derived with high accuracy from raw texts. Hybrid models can be used to extract verifiable facts at scale to build controllable and interpretable systems, for grounding in human-robot interaction (HRI) systems, support logical reasoning, or used in extremely low resource settings. This tutorial will provide an overview of these semantic representations, the computational models that are trained on them, as well as the practical applications built with these representations, including future directions. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/20meaning-representations-for-natural-languages-design-models-and-applications-240618162649-7129583c-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> COLING-LREC&#39;2024 Tutorial &quot;Meaning Representations for Natural Languages: Design, Models and Applications&quot; Instructors: Julia Bonn, Jeffrey Flanigan, Jan Haji, Ishan Jindal, Yunyao Li and Nianwen Xue Abstract: This tutorial introduces a research area that has the potential to create linguistic resources and build computational models that provide critical components for interpretable and controllable NLP systems. While large language models have shown remarkable ability to generate fluent text, the blackbox nature of these models makes it difficult to know where to tweak these models to fix errors, at least for now. For instance, LLMs are known to hallucinate and there is no mechanism in these models to only provide factually correct answers. Addressing this issue requires that first of all the models have access to a body of verifiable facts, and then use it effectively. Interpretability and controllability in NLP systems are critical in high-stake applications such as the medical domain. There has been a steady accumulation of semantically annotated, increasingly richer resources, which can now be derived with high accuracy from raw texts. Hybrid models can be used to extract verifiable facts at scale to build controllable and interpretable systems, for grounding in human-robot interaction (HRI) systems, support logical reasoning, or used in extremely low resource settings. This tutorial will provide an overview of these semantic representations, the computational models that are trained on them, as well as the practical applications built with these representations, including future directions.
Meaning Representations for-Natural Languages Design, Models, and Applications.pdf from Yunyao Li
]]> 52 0 https://cdn.slidesharecdn.com/ss_thumbnails/20meaning-representations-for-natural-languages-design-models-and-applications-240618162649-7129583c-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 The Role of Patterns in the Era of Large Language Models /slideshow/the-role-of-patterns-in-the-era-of-large-language-models/264351235 pandl-emnlp-december-2023-231206072657-815069fe
際際滷s for my keynote at PAN-DL Workshop (Pattern-based Approaches to NLP in the Age of Deep Learning) at EMNLP'2023 (December. 6, 2023). In this talk, I share our initial learnings from constructing, growing and serving large knowledge graphs]]>
際際滷s for my keynote at PAN-DL Workshop (Pattern-based Approaches to NLP in the Age of Deep Learning) at EMNLP'2023 (December. 6, 2023). In this talk, I share our initial learnings from constructing, growing and serving large knowledge graphs]]> Wed, 06 Dec 2023 07:26:57 GMT /slideshow/the-role-of-patterns-in-the-era-of-large-language-models/264351235 YunyaoLi@slideshare.net(YunyaoLi) The Role of Patterns in the Era of Large Language Models YunyaoLi 際際滷s for my keynote at PAN-DL Workshop (Pattern-based Approaches to NLP in the Age of Deep Learning) at EMNLP'2023 (December. 6, 2023). In this talk, I share our initial learnings from constructing, growing and serving large knowledge graphs <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/pandl-emnlp-december-2023-231206072657-815069fe-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> 際際滷s for my keynote at PAN-DL Workshop (Pattern-based Approaches to NLP in the Age of Deep Learning) at EMNLP&#39;2023 (December. 6, 2023). In this talk, I share our initial learnings from constructing, growing and serving large knowledge graphs
The Role of Patterns in the Era of Large Language Models from Yunyao Li
]]> 284 0 https://cdn.slidesharecdn.com/ss_thumbnails/pandl-emnlp-december-2023-231206072657-815069fe-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Building, Growing and Serving Large Knowledge Graphs with Human-in-the-Loop /slideshow/building-growing-and-serving-large-knowledge-graphs-with-humanintheloop/258485723 hilda-keynote-june-18-2023-v1-230618213135-4a78f89d
Keynote talk at HILDA'2023 at SIGMOD on June 18, 2023. Abstract: The ability to build large-scale knowledge bases that capture and extend the implicit knowledge of human experts is the foundation for many AI systems. We use an ontology-driven approach for the building, growing and serving of such knowledge bases. This approach relies on several well-known building blocks: document conversion, natural language processing, entity resolution, data transformation and fusion. In this talk, I will discuss wide range of real-world challenges related to the building of these blocks and present our work to address these challenges via better human-machine cooperation. ]]>
Keynote talk at HILDA'2023 at SIGMOD on June 18, 2023. Abstract: The ability to build large-scale knowledge bases that capture and extend the implicit knowledge of human experts is the foundation for many AI systems. We use an ontology-driven approach for the building, growing and serving of such knowledge bases. This approach relies on several well-known building blocks: document conversion, natural language processing, entity resolution, data transformation and fusion. In this talk, I will discuss wide range of real-world challenges related to the building of these blocks and present our work to address these challenges via better human-machine cooperation. ]]> Sun, 18 Jun 2023 21:31:35 GMT /slideshow/building-growing-and-serving-large-knowledge-graphs-with-humanintheloop/258485723 YunyaoLi@slideshare.net(YunyaoLi) Building, Growing and Serving Large Knowledge Graphs with Human-in-the-Loop YunyaoLi Keynote talk at HILDA'2023 at SIGMOD on June 18, 2023. Abstract: The ability to build large-scale knowledge bases that capture and extend the implicit knowledge of human experts is the foundation for many AI systems. We use an ontology-driven approach for the building, growing and serving of such knowledge bases. This approach relies on several well-known building blocks: document conversion, natural language processing, entity resolution, data transformation and fusion. In this talk, I will discuss wide range of real-world challenges related to the building of these blocks and present our work to address these challenges via better human-machine cooperation. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/hilda-keynote-june-18-2023-v1-230618213135-4a78f89d-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Keynote talk at HILDA&#39;2023 at SIGMOD on June 18, 2023. Abstract: The ability to build large-scale knowledge bases that capture and extend the implicit knowledge of human experts is the foundation for many AI systems. We use an ontology-driven approach for the building, growing and serving of such knowledge bases. This approach relies on several well-known building blocks: document conversion, natural language processing, entity resolution, data transformation and fusion. In this talk, I will discuss wide range of real-world challenges related to the building of these blocks and present our work to address these challenges via better human-machine cooperation.
Building, Growing and Serving Large Knowledge Graphs with Human-in-the-Loop from Yunyao Li
]]> 378 0 https://cdn.slidesharecdn.com/ss_thumbnails/hilda-keynote-june-18-2023-v1-230618213135-4a78f89d-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Meaning Representations for Natural Languages: Design, Models and Applications /slideshow/meaning-representations-for-natural-languages-design-models-and-applications/255676647 2022emnlpt1meaningrepresentationsfornaturallanguages-230203042616-a19065d6
EMNLP'2022 Tutorial "Meaning Representations for Natural Languages: Design, Models and Applications" Instructors: Jeffrey Flanigan, Ishan Jindal, Yunyao Li, Tim OGorman, Martha Palmer Abstract: We propose a cutting-edge tutorial that reviews the design of common meaning representations, SoTA models for predicting meaning representations, and the applications of meaning representations in a wide range of downstream NLP tasks and real-world applications. Reporting by a diverse team of NLP researchers from academia and industry with extensive experience in designing, building and using meaning representations, our tutorial has three components: (1) an introduction to common meaning representations, including basic concepts and design challenges; (2) a review of SoTA methods on building models for meaning representations; and (3) an overview of applications of meaning representations in downstream NLP tasks and real-world applications. We will also present qualitative comparisons of common meaning representations and a quantitative study on how their differences impact model performance. Finally, we will share best practices in choosing the right meaning representation for downstream tasks.]]>
EMNLP'2022 Tutorial "Meaning Representations for Natural Languages: Design, Models and Applications" Instructors: Jeffrey Flanigan, Ishan Jindal, Yunyao Li, Tim OGorman, Martha Palmer Abstract: We propose a cutting-edge tutorial that reviews the design of common meaning representations, SoTA models for predicting meaning representations, and the applications of meaning representations in a wide range of downstream NLP tasks and real-world applications. Reporting by a diverse team of NLP researchers from academia and industry with extensive experience in designing, building and using meaning representations, our tutorial has three components: (1) an introduction to common meaning representations, including basic concepts and design challenges; (2) a review of SoTA methods on building models for meaning representations; and (3) an overview of applications of meaning representations in downstream NLP tasks and real-world applications. We will also present qualitative comparisons of common meaning representations and a quantitative study on how their differences impact model performance. Finally, we will share best practices in choosing the right meaning representation for downstream tasks.]]> Fri, 03 Feb 2023 04:26:16 GMT /slideshow/meaning-representations-for-natural-languages-design-models-and-applications/255676647 YunyaoLi@slideshare.net(YunyaoLi) Meaning Representations for Natural Languages: Design, Models and Applications YunyaoLi EMNLP'2022 Tutorial "Meaning Representations for Natural Languages: Design, Models and Applications" Instructors: Jeffrey Flanigan, Ishan Jindal, Yunyao Li, Tim OGorman, Martha Palmer Abstract: We propose a cutting-edge tutorial that reviews the design of common meaning representations, SoTA models for predicting meaning representations, and the applications of meaning representations in a wide range of downstream NLP tasks and real-world applications. Reporting by a diverse team of NLP researchers from academia and industry with extensive experience in designing, building and using meaning representations, our tutorial has three components: (1) an introduction to common meaning representations, including basic concepts and design challenges; (2) a review of SoTA methods on building models for meaning representations; and (3) an overview of applications of meaning representations in downstream NLP tasks and real-world applications. We will also present qualitative comparisons of common meaning representations and a quantitative study on how their differences impact model performance. Finally, we will share best practices in choosing the right meaning representation for downstream tasks. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/2022emnlpt1meaningrepresentationsfornaturallanguages-230203042616-a19065d6-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> EMNLP&#39;2022 Tutorial &quot;Meaning Representations for Natural Languages: Design, Models and Applications&quot; Instructors: Jeffrey Flanigan, Ishan Jindal, Yunyao Li, Tim OGorman, Martha Palmer Abstract: We propose a cutting-edge tutorial that reviews the design of common meaning representations, SoTA models for predicting meaning representations, and the applications of meaning representations in a wide range of downstream NLP tasks and real-world applications. Reporting by a diverse team of NLP researchers from academia and industry with extensive experience in designing, building and using meaning representations, our tutorial has three components: (1) an introduction to common meaning representations, including basic concepts and design challenges; (2) a review of SoTA methods on building models for meaning representations; and (3) an overview of applications of meaning representations in downstream NLP tasks and real-world applications. We will also present qualitative comparisons of common meaning representations and a quantitative study on how their differences impact model performance. Finally, we will share best practices in choosing the right meaning representation for downstream tasks.
Meaning Representations for Natural Languages: Design, Models and Applications from Yunyao Li
]]> 611 0 https://cdn.slidesharecdn.com/ss_thumbnails/2022emnlpt1meaningrepresentationsfornaturallanguages-230203042616-a19065d6-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Taming the Wild West of NLP /slideshow/taming-the-wild-west-of-nlp/250641512 tamingwildwestofnlp-211112222756
Natural language understanding is a fundamental task in artificial intelligence. English understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Discovery. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in the past few years in addressing these challenges. We will also showcase how universal semantic representation of natural languages can enable cross-lingual information extraction in concrete domains (e.g. compliance) and show ongoing efforts towards seamless scaling existing NLP capabilities across languages with minimal efforts. ]]>
Natural language understanding is a fundamental task in artificial intelligence. English understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Discovery. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in the past few years in addressing these challenges. We will also showcase how universal semantic representation of natural languages can enable cross-lingual information extraction in concrete domains (e.g. compliance) and show ongoing efforts towards seamless scaling existing NLP capabilities across languages with minimal efforts. ]]> Fri, 12 Nov 2021 22:27:55 GMT /slideshow/taming-the-wild-west-of-nlp/250641512 YunyaoLi@slideshare.net(YunyaoLi) Taming the Wild West of NLP YunyaoLi Natural language understanding is a fundamental task in artificial intelligence. English understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Discovery. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in the past few years in addressing these challenges. We will also showcase how universal semantic representation of natural languages can enable cross-lingual information extraction in concrete domains (e.g. compliance) and show ongoing efforts towards seamless scaling existing NLP capabilities across languages with minimal efforts. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/tamingwildwestofnlp-211112222756-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Natural language understanding is a fundamental task in artificial intelligence. English understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Discovery. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in the past few years in addressing these challenges. We will also showcase how universal semantic representation of natural languages can enable cross-lingual information extraction in concrete domains (e.g. compliance) and show ongoing efforts towards seamless scaling existing NLP capabilities across languages with minimal efforts.
Taming the Wild West of NLP from Yunyao Li
]]> 450 0 https://cdn.slidesharecdn.com/ss_thumbnails/tamingwildwestofnlp-211112222756-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Towards Deep Table Understanding /slideshow/towards-deep-table-understanding/250011244 towardsdeeptableundersanding-210819161054
Invited talk at Document Intelligence workshop at KDD'2021. Harvesting information from complex documents such as in financial reports and scientific publications is critical to building AI applications for business and research. Such documents are often in PDF format with critical facts and data conveyed in table and graphs. Extracting such information is essential to extract insights from these documents. In IBM Research, we have a rich agenda in this area that we call Deep Document Understanding. In this talk, I will focus on our research on Deep Table Understanding extracting and understanding tables from PDF documents. I will introduce key challenges in table extraction and understanding and how we address such challenges, from how to acquire data at scale to enable deep neural network models to how to build, customize and evaluate such models. I will also describe how our work enables real-world use cases in domains such as finance and life science. Finally, I will briefly present TableQA, an important downstream task enabled by Deep Table Understanding.]]>
Invited talk at Document Intelligence workshop at KDD'2021. Harvesting information from complex documents such as in financial reports and scientific publications is critical to building AI applications for business and research. Such documents are often in PDF format with critical facts and data conveyed in table and graphs. Extracting such information is essential to extract insights from these documents. In IBM Research, we have a rich agenda in this area that we call Deep Document Understanding. In this talk, I will focus on our research on Deep Table Understanding extracting and understanding tables from PDF documents. I will introduce key challenges in table extraction and understanding and how we address such challenges, from how to acquire data at scale to enable deep neural network models to how to build, customize and evaluate such models. I will also describe how our work enables real-world use cases in domains such as finance and life science. Finally, I will briefly present TableQA, an important downstream task enabled by Deep Table Understanding.]]> Thu, 19 Aug 2021 16:10:53 GMT /slideshow/towards-deep-table-understanding/250011244 YunyaoLi@slideshare.net(YunyaoLi) Towards Deep Table Understanding YunyaoLi Invited talk at Document Intelligence workshop at KDD'2021. Harvesting information from complex documents such as in financial reports and scientific publications is critical to building AI applications for business and research. Such documents are often in PDF format with critical facts and data conveyed in table and graphs. Extracting such information is essential to extract insights from these documents. In IBM Research, we have a rich agenda in this area that we call Deep Document Understanding. In this talk, I will focus on our research on Deep Table Understanding extracting and understanding tables from PDF documents. I will introduce key challenges in table extraction and understanding and how we address such challenges, from how to acquire data at scale to enable deep neural network models to how to build, customize and evaluate such models. I will also describe how our work enables real-world use cases in domains such as finance and life science. Finally, I will briefly present TableQA, an important downstream task enabled by Deep Table Understanding. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/towardsdeeptableundersanding-210819161054-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Invited talk at Document Intelligence workshop at KDD&#39;2021. Harvesting information from complex documents such as in financial reports and scientific publications is critical to building AI applications for business and research. Such documents are often in PDF format with critical facts and data conveyed in table and graphs. Extracting such information is essential to extract insights from these documents. In IBM Research, we have a rich agenda in this area that we call Deep Document Understanding. In this talk, I will focus on our research on Deep Table Understanding extracting and understanding tables from PDF documents. I will introduce key challenges in table extraction and understanding and how we address such challenges, from how to acquire data at scale to enable deep neural network models to how to build, customize and evaluate such models. I will also describe how our work enables real-world use cases in domains such as finance and life science. Finally, I will briefly present TableQA, an important downstream task enabled by Deep Table Understanding.
Towards Deep Table Understanding from Yunyao Li
]]> 135 0 https://cdn.slidesharecdn.com/ss_thumbnails/towardsdeeptableundersanding-210819161054-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Explainability for Natural Language Processing /slideshow/explainability-for-natural-language-processing-249992241/249992241 kdd2021-xai-tutorial-final-210817001151
Final deck for our popular tutorial on "Explainability for Natural Language Processing" at KDD'2021. See links below for downloadable version (with higher resolution) and recording of the live tutorial. Title: Explainability for Natural Language Processing Presenter: Marina Danilevsky, Shipi Dhanorkar, Yunyao Li and Lucian Popa and Kun Qian and Anbang Xu Website: http://xainlp.github.io/ Recording: https://www.youtube.com/watch?v=PvKOSYGclPk&t=2s Downloadable version with higher resolution: https://drive.google.com/file/d/1_gt_cS9nP9rcZOn4dcmxc2CErxrHW9CU/view?usp=sharing @article{kdd2021xaitutorial, title={Explainability for Natural Language Processing}, author= {Marina Danilevsky, Shipi Dhanorkar and Yunyao Li and Lucian Popa and Kun Qian and Anbang Xu}, journal={KDD}, year={2021} } Abstract: This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable.Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability inNLP. Then, we will discuss explainability for NLP tasks and reporton a systematic literature review of the state-of-the-art literaturein AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data.]]>
Final deck for our popular tutorial on "Explainability for Natural Language Processing" at KDD'2021. See links below for downloadable version (with higher resolution) and recording of the live tutorial. Title: Explainability for Natural Language Processing Presenter: Marina Danilevsky, Shipi Dhanorkar, Yunyao Li and Lucian Popa and Kun Qian and Anbang Xu Website: http://xainlp.github.io/ Recording: https://www.youtube.com/watch?v=PvKOSYGclPk&t=2s Downloadable version with higher resolution: https://drive.google.com/file/d/1_gt_cS9nP9rcZOn4dcmxc2CErxrHW9CU/view?usp=sharing @article{kdd2021xaitutorial, title={Explainability for Natural Language Processing}, author= {Marina Danilevsky, Shipi Dhanorkar and Yunyao Li and Lucian Popa and Kun Qian and Anbang Xu}, journal={KDD}, year={2021} } Abstract: This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable.Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability inNLP. Then, we will discuss explainability for NLP tasks and reporton a systematic literature review of the state-of-the-art literaturein AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data.]]> Tue, 17 Aug 2021 00:11:50 GMT /slideshow/explainability-for-natural-language-processing-249992241/249992241 YunyaoLi@slideshare.net(YunyaoLi) Explainability for Natural Language Processing YunyaoLi Final deck for our popular tutorial on "Explainability for Natural Language Processing" at KDD'2021. See links below for downloadable version (with higher resolution) and recording of the live tutorial. Title: Explainability for Natural Language Processing Presenter: Marina Danilevsky, Shipi Dhanorkar, Yunyao Li and Lucian Popa and Kun Qian and Anbang Xu Website: http://xainlp.github.io/ Recording: https://www.youtube.com/watch?v=PvKOSYGclPk&t=2s Downloadable version with higher resolution: https://drive.google.com/file/d/1_gt_cS9nP9rcZOn4dcmxc2CErxrHW9CU/view?usp=sharing @article{kdd2021xaitutorial, title={Explainability for Natural Language Processing}, author= {Marina Danilevsky, Shipi Dhanorkar and Yunyao Li and Lucian Popa and Kun Qian and Anbang Xu}, journal={KDD}, year={2021} } Abstract: This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable.Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability inNLP. Then, we will discuss explainability for NLP tasks and reporton a systematic literature review of the state-of-the-art literaturein AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/kdd2021-xai-tutorial-final-210817001151-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Final deck for our popular tutorial on &quot;Explainability for Natural Language Processing&quot; at KDD&#39;2021. See links below for downloadable version (with higher resolution) and recording of the live tutorial. Title: Explainability for Natural Language Processing Presenter: Marina Danilevsky, Shipi Dhanorkar, Yunyao Li and Lucian Popa and Kun Qian and Anbang Xu Website: http://xainlp.github.io/ Recording: https://www.youtube.com/watch?v=PvKOSYGclPk&amp;t=2s Downloadable version with higher resolution: https://drive.google.com/file/d/1_gt_cS9nP9rcZOn4dcmxc2CErxrHW9CU/view?usp=sharing @article{kdd2021xaitutorial, title={Explainability for Natural Language Processing}, author= {Marina Danilevsky, Shipi Dhanorkar and Yunyao Li and Lucian Popa and Kun Qian and Anbang Xu}, journal={KDD}, year={2021} } Abstract: This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable.Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability inNLP. Then, we will discuss explainability for NLP tasks and reporton a systematic literature review of the state-of-the-art literaturein AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data.
Explainability for Natural Language Processing from Yunyao Li
]]> 2734 0 https://cdn.slidesharecdn.com/ss_thumbnails/kdd2021-xai-tutorial-final-210817001151-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Explainability for Natural Language Processing /YunyaoLi/explainability-for-natural-language-processing-249912819 finalv2-210803185838
NOTE: Please check out the final version here with small but important updates and links to downloadable version and recording: /YunyaoLi/explainability-for-natural-language-processing-249992241 Updated version on our popular tutorial on "Explainability for Natural Language Processing" as a tutorial at KDD'2021. Title: Explainability for Natural Language Processing @article{kdd2021xaitutorial, title={Explainability for Natural Language Processing}, author= {Marina Danilevsky, Dhanorkar, Shipi and Li, Yunyao and Lucian Popa and Kun Qian and Anbang Xu}, journal={KDD}, year={2021} } Presenter: Marina Danilevsky, Dhanorkar, Shipi and Li, Yunyao and Lucian Popa and Kun Qian and Anbang Xu Website: http://xainlp.github.io/ Abstract: This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable.Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability inNLP. Then, we will discuss explainability for NLP tasks and reporton a systematic literature review of the state-of-the-art literaturein AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data.]]>
NOTE: Please check out the final version here with small but important updates and links to downloadable version and recording: /YunyaoLi/explainability-for-natural-language-processing-249992241 Updated version on our popular tutorial on "Explainability for Natural Language Processing" as a tutorial at KDD'2021. Title: Explainability for Natural Language Processing @article{kdd2021xaitutorial, title={Explainability for Natural Language Processing}, author= {Marina Danilevsky, Dhanorkar, Shipi and Li, Yunyao and Lucian Popa and Kun Qian and Anbang Xu}, journal={KDD}, year={2021} } Presenter: Marina Danilevsky, Dhanorkar, Shipi and Li, Yunyao and Lucian Popa and Kun Qian and Anbang Xu Website: http://xainlp.github.io/ Abstract: This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable.Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability inNLP. Then, we will discuss explainability for NLP tasks and reporton a systematic literature review of the state-of-the-art literaturein AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data.]]> Tue, 03 Aug 2021 18:58:38 GMT /YunyaoLi/explainability-for-natural-language-processing-249912819 YunyaoLi@slideshare.net(YunyaoLi) Explainability for Natural Language Processing YunyaoLi NOTE: Please check out the final version here with small but important updates and links to downloadable version and recording: /YunyaoLi/explainability-for-natural-language-processing-249992241 Updated version on our popular tutorial on "Explainability for Natural Language Processing" as a tutorial at KDD'2021. Title: Explainability for Natural Language Processing @article{kdd2021xaitutorial, title={Explainability for Natural Language Processing}, author= {Marina Danilevsky, Dhanorkar, Shipi and Li, Yunyao and Lucian Popa and Kun Qian and Anbang Xu}, journal={KDD}, year={2021} } Presenter: Marina Danilevsky, Dhanorkar, Shipi and Li, Yunyao and Lucian Popa and Kun Qian and Anbang Xu Website: http://xainlp.github.io/ Abstract: This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable.Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability inNLP. Then, we will discuss explainability for NLP tasks and reporton a systematic literature review of the state-of-the-art literaturein AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/finalv2-210803185838-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> NOTE: Please check out the final version here with small but important updates and links to downloadable version and recording: /YunyaoLi/explainability-for-natural-language-processing-249992241 Updated version on our popular tutorial on &quot;Explainability for Natural Language Processing&quot; as a tutorial at KDD&#39;2021. Title: Explainability for Natural Language Processing @article{kdd2021xaitutorial, title={Explainability for Natural Language Processing}, author= {Marina Danilevsky, Dhanorkar, Shipi and Li, Yunyao and Lucian Popa and Kun Qian and Anbang Xu}, journal={KDD}, year={2021} } Presenter: Marina Danilevsky, Dhanorkar, Shipi and Li, Yunyao and Lucian Popa and Kun Qian and Anbang Xu Website: http://xainlp.github.io/ Abstract: This lecture-style tutorial, which mixes in an interactive literature browsing component, is intended for the many researchers and practitioners working with text data and on applications of natural language processing (NLP) in data science and knowledge discovery. The focus of the tutorial is on the issues of transparency and interpretability as they relate to building models for text and their applications to knowledge discovery. As black-box models have gained popularity for a broad range of tasks in recent years, both the research and industry communities have begun developing new techniques to render them more transparent and interpretable.Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP/knowledge management researchers, our tutorial has two components: an introduction to explainable AI (XAI) in the NLP domain and a review of the state-of-the-art research; and findings from a qualitative interview study of individuals working on real-world NLP projects as they are applied to various knowledge extraction and discovery at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability inNLP. Then, we will discuss explainability for NLP tasks and reporton a systematic literature review of the state-of-the-art literaturein AI, NLP and HCI conferences. The second component reports on our qualitative interview study, which identifies practical challenges and concerns that arise in real-world development projects that require the modeling and understanding of text data.
Explainability for Natural Language Processing from Yunyao Li
]]> 3222 0 https://cdn.slidesharecdn.com/ss_thumbnails/finalv2-210803185838-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Human in the Loop AI for Building Knowledge Bases /YunyaoLi/human-in-the-loop-ai-for-building-knowledge-bases buildingdomainspecifickb-210522044055
The ability to build large-scale domain-specific knowledge bases that capture and extend the implicit knowledge of human experts is the foundation for many AI systems. We use an ontology-driven approach for the creation, representation and consumption of such domain-specific knowledge bases. This approach relies on several well-known building blocks: natural language processing, entity resolution, data transformation and fusion. I will present several human-in-the-loop work that target domain experts (rather than programmers) to extract the domain knowledge from the human expert and map it into the "right" models or algorithms. I will also share successful use cases in several domains, including Compliance, Finance, and Healthcare: by using these tools we can match the level of accuracy achieved by manual efforts, but at a significantly lower cost and much higher scale and automation. ]]>
The ability to build large-scale domain-specific knowledge bases that capture and extend the implicit knowledge of human experts is the foundation for many AI systems. We use an ontology-driven approach for the creation, representation and consumption of such domain-specific knowledge bases. This approach relies on several well-known building blocks: natural language processing, entity resolution, data transformation and fusion. I will present several human-in-the-loop work that target domain experts (rather than programmers) to extract the domain knowledge from the human expert and map it into the "right" models or algorithms. I will also share successful use cases in several domains, including Compliance, Finance, and Healthcare: by using these tools we can match the level of accuracy achieved by manual efforts, but at a significantly lower cost and much higher scale and automation. ]]> Sat, 22 May 2021 04:40:55 GMT /YunyaoLi/human-in-the-loop-ai-for-building-knowledge-bases YunyaoLi@slideshare.net(YunyaoLi) Human in the Loop AI for Building Knowledge Bases YunyaoLi The ability to build large-scale domain-specific knowledge bases that capture and extend the implicit knowledge of human experts is the foundation for many AI systems. We use an ontology-driven approach for the creation, representation and consumption of such domain-specific knowledge bases. This approach relies on several well-known building blocks: natural language processing, entity resolution, data transformation and fusion. I will present several human-in-the-loop work that target domain experts (rather than programmers) to extract the domain knowledge from the human expert and map it into the "right" models or algorithms. I will also share successful use cases in several domains, including Compliance, Finance, and Healthcare: by using these tools we can match the level of accuracy achieved by manual efforts, but at a significantly lower cost and much higher scale and automation. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/buildingdomainspecifickb-210522044055-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> The ability to build large-scale domain-specific knowledge bases that capture and extend the implicit knowledge of human experts is the foundation for many AI systems. We use an ontology-driven approach for the creation, representation and consumption of such domain-specific knowledge bases. This approach relies on several well-known building blocks: natural language processing, entity resolution, data transformation and fusion. I will present several human-in-the-loop work that target domain experts (rather than programmers) to extract the domain knowledge from the human expert and map it into the &quot;right&quot; models or algorithms. I will also share successful use cases in several domains, including Compliance, Finance, and Healthcare: by using these tools we can match the level of accuracy achieved by manual efforts, but at a significantly lower cost and much higher scale and automation.
Human in the Loop AI for Building Knowledge Bases from Yunyao Li
]]> 303 0 https://cdn.slidesharecdn.com/ss_thumbnails/buildingdomainspecifickb-210522044055-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Towards Universal Language Understanding /slideshow/towards-universal-language-understanding-244833734/244833734 2021-210320180026
際際滷s for talk given at Women in Engineering on March 20, 2021. Abstract: Natural language understanding is a fundamental task in artificial intelligence. English understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Discovery. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in the past few years in addressing these challenges. We will also showcase how universal semantic representation of natural languages can enable cross-lingual information extraction in concrete domains (e.g. compliance) and show ongoing efforts towards seamless scaling existing NLP capabilities across languages with minimal efforts.]]>
際際滷s for talk given at Women in Engineering on March 20, 2021. Abstract: Natural language understanding is a fundamental task in artificial intelligence. English understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Discovery. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in the past few years in addressing these challenges. We will also showcase how universal semantic representation of natural languages can enable cross-lingual information extraction in concrete domains (e.g. compliance) and show ongoing efforts towards seamless scaling existing NLP capabilities across languages with minimal efforts.]]> Sat, 20 Mar 2021 18:00:25 GMT /slideshow/towards-universal-language-understanding-244833734/244833734 YunyaoLi@slideshare.net(YunyaoLi) Towards Universal Language Understanding YunyaoLi 際際滷s for talk given at Women in Engineering on March 20, 2021. Abstract: Natural language understanding is a fundamental task in artificial intelligence. English understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Discovery. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in the past few years in addressing these challenges. We will also showcase how universal semantic representation of natural languages can enable cross-lingual information extraction in concrete domains (e.g. compliance) and show ongoing efforts towards seamless scaling existing NLP capabilities across languages with minimal efforts. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/2021-210320180026-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> 際際滷s for talk given at Women in Engineering on March 20, 2021. Abstract: Natural language understanding is a fundamental task in artificial intelligence. English understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Discovery. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in the past few years in addressing these challenges. We will also showcase how universal semantic representation of natural languages can enable cross-lingual information extraction in concrete domains (e.g. compliance) and show ongoing efforts towards seamless scaling existing NLP capabilities across languages with minimal efforts.
Towards Universal Language Understanding from Yunyao Li
]]> 228 0 https://cdn.slidesharecdn.com/ss_thumbnails/2021-210320180026-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Explainability for Natural Language Processing /slideshow/explainability-for-natural-language-processing/239767498 aacl2020-xai-tutorial-final-201205000527
Tutorial at AACL'2020 (http://www.aacl2020.org/program/tutorials/#t4-explainability-for-natural-language-processing). More recent version: /YunyaoLi/explainability-for-natural-language-processing-249912819 Title: Explainability for Natural Language Processing @article{aacl2020xaitutorial, title={Explainability for Natural Language Processing}, author= {Dhanorkar, Shipi and Li, Yunyao and Popa, Lucian and Qian, Kun and Wolf, Christine T and Xu, Anbang}, journal={AACL-IJCNLP 2020}, year={2020} Presenter: Shipi Dhanorkar, Christine Wolf, Kun Qian, Anbang Xu, Lucian Popa and Yunyao Li Video: https://www.youtube.com/watch?v=3tnrGe_JA0s&feature=youtu.be Abstract: We propose a cutting-edge tutorial that investigates the issues of transparency and interpretability as they relate to NLP. Both the research community and industry have been developing new techniques to render black-box NLP models more transparent and interpretable. Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP researchers, our tutorial has two components: an introduction to explainable AI (XAI) and a review of the state-of-the-art for explainability research in NLP; and findings from a qualitative interview study of individuals working on real-world NLP projects at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability in NLP. Then, we will discuss explainability for NLP tasks and report on a systematic literature review of the state-of-the-art literature in AI, NLP, and HCI conferences. The second component reports on our qualitative interview study which identifies practical challenges and concerns that arise in real-world development projects which include NLP.]]>
Tutorial at AACL'2020 (http://www.aacl2020.org/program/tutorials/#t4-explainability-for-natural-language-processing). More recent version: /YunyaoLi/explainability-for-natural-language-processing-249912819 Title: Explainability for Natural Language Processing @article{aacl2020xaitutorial, title={Explainability for Natural Language Processing}, author= {Dhanorkar, Shipi and Li, Yunyao and Popa, Lucian and Qian, Kun and Wolf, Christine T and Xu, Anbang}, journal={AACL-IJCNLP 2020}, year={2020} Presenter: Shipi Dhanorkar, Christine Wolf, Kun Qian, Anbang Xu, Lucian Popa and Yunyao Li Video: https://www.youtube.com/watch?v=3tnrGe_JA0s&feature=youtu.be Abstract: We propose a cutting-edge tutorial that investigates the issues of transparency and interpretability as they relate to NLP. Both the research community and industry have been developing new techniques to render black-box NLP models more transparent and interpretable. Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP researchers, our tutorial has two components: an introduction to explainable AI (XAI) and a review of the state-of-the-art for explainability research in NLP; and findings from a qualitative interview study of individuals working on real-world NLP projects at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability in NLP. Then, we will discuss explainability for NLP tasks and report on a systematic literature review of the state-of-the-art literature in AI, NLP, and HCI conferences. The second component reports on our qualitative interview study which identifies practical challenges and concerns that arise in real-world development projects which include NLP.]]> Sat, 05 Dec 2020 00:05:26 GMT /slideshow/explainability-for-natural-language-processing/239767498 YunyaoLi@slideshare.net(YunyaoLi) Explainability for Natural Language Processing YunyaoLi Tutorial at AACL'2020 (http://www.aacl2020.org/program/tutorials/#t4-explainability-for-natural-language-processing). More recent version: /YunyaoLi/explainability-for-natural-language-processing-249912819 Title: Explainability for Natural Language Processing @article{aacl2020xaitutorial, title={Explainability for Natural Language Processing}, author= {Dhanorkar, Shipi and Li, Yunyao and Popa, Lucian and Qian, Kun and Wolf, Christine T and Xu, Anbang}, journal={AACL-IJCNLP 2020}, year={2020} Presenter: Shipi Dhanorkar, Christine Wolf, Kun Qian, Anbang Xu, Lucian Popa and Yunyao Li Video: https://www.youtube.com/watch?v=3tnrGe_JA0s&feature=youtu.be Abstract: We propose a cutting-edge tutorial that investigates the issues of transparency and interpretability as they relate to NLP. Both the research community and industry have been developing new techniques to render black-box NLP models more transparent and interpretable. Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP researchers, our tutorial has two components: an introduction to explainable AI (XAI) and a review of the state-of-the-art for explainability research in NLP; and findings from a qualitative interview study of individuals working on real-world NLP projects at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability in NLP. Then, we will discuss explainability for NLP tasks and report on a systematic literature review of the state-of-the-art literature in AI, NLP, and HCI conferences. The second component reports on our qualitative interview study which identifies practical challenges and concerns that arise in real-world development projects which include NLP. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/aacl2020-xai-tutorial-final-201205000527-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Tutorial at AACL&#39;2020 (http://www.aacl2020.org/program/tutorials/#t4-explainability-for-natural-language-processing). More recent version: /YunyaoLi/explainability-for-natural-language-processing-249912819 Title: Explainability for Natural Language Processing @article{aacl2020xaitutorial, title={Explainability for Natural Language Processing}, author= {Dhanorkar, Shipi and Li, Yunyao and Popa, Lucian and Qian, Kun and Wolf, Christine T and Xu, Anbang}, journal={AACL-IJCNLP 2020}, year={2020} Presenter: Shipi Dhanorkar, Christine Wolf, Kun Qian, Anbang Xu, Lucian Popa and Yunyao Li Video: https://www.youtube.com/watch?v=3tnrGe_JA0s&amp;feature=youtu.be Abstract: We propose a cutting-edge tutorial that investigates the issues of transparency and interpretability as they relate to NLP. Both the research community and industry have been developing new techniques to render black-box NLP models more transparent and interpretable. Reporting from an interdisciplinary team of social science, human-computer interaction (HCI), and NLP researchers, our tutorial has two components: an introduction to explainable AI (XAI) and a review of the state-of-the-art for explainability research in NLP; and findings from a qualitative interview study of individuals working on real-world NLP projects at a large, multinational technology and consulting corporation. The first component will introduce core concepts related to explainability in NLP. Then, we will discuss explainability for NLP tasks and report on a systematic literature review of the state-of-the-art literature in AI, NLP, and HCI conferences. The second component reports on our qualitative interview study which identifies practical challenges and concerns that arise in real-world development projects which include NLP.
Explainability for Natural Language Processing from Yunyao Li
]]> 4393 0 https://cdn.slidesharecdn.com/ss_thumbnails/aacl2020-xai-tutorial-final-201205000527-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Towards Universal Language Understanding (2020 version) /slideshow/towards-universal-language-understanding-2020-version/239691001 towardsuniversallanguageunderstanding-2-201202181619
Keynote talk given at Pacific Asia Conference on Language, Information and Computation (PACLIC 34) on Pacific Asia Conference on Language, Information and Computation (PACLIC 34) on October 24, 2020. Title: Towards Universal Natural Language Understanding Abstract: Understanding the semantics of the natural language is a fundamental task in artificial intelligence. English semantic understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Compare and Comply. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in addressing these challenges in the past few years to provide the same unified semantic representation across languages. We will also showcase how such universal semantic understanding of natural languages can enable cross-lingual information extraction in concrete domains (e.g. insurance and compliance) and show promise towards seamless scaling existing NLP capabilities across languages with minimal efforts.]]>
Keynote talk given at Pacific Asia Conference on Language, Information and Computation (PACLIC 34) on Pacific Asia Conference on Language, Information and Computation (PACLIC 34) on October 24, 2020. Title: Towards Universal Natural Language Understanding Abstract: Understanding the semantics of the natural language is a fundamental task in artificial intelligence. English semantic understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Compare and Comply. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in addressing these challenges in the past few years to provide the same unified semantic representation across languages. We will also showcase how such universal semantic understanding of natural languages can enable cross-lingual information extraction in concrete domains (e.g. insurance and compliance) and show promise towards seamless scaling existing NLP capabilities across languages with minimal efforts.]]> Wed, 02 Dec 2020 18:16:18 GMT /slideshow/towards-universal-language-understanding-2020-version/239691001 YunyaoLi@slideshare.net(YunyaoLi) Towards Universal Language Understanding (2020 version) YunyaoLi Keynote talk given at Pacific Asia Conference on Language, Information and Computation (PACLIC 34) on Pacific Asia Conference on Language, Information and Computation (PACLIC 34) on October 24, 2020. Title: Towards Universal Natural Language Understanding Abstract: Understanding the semantics of the natural language is a fundamental task in artificial intelligence. English semantic understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Compare and Comply. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in addressing these challenges in the past few years to provide the same unified semantic representation across languages. We will also showcase how such universal semantic understanding of natural languages can enable cross-lingual information extraction in concrete domains (e.g. insurance and compliance) and show promise towards seamless scaling existing NLP capabilities across languages with minimal efforts. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/towardsuniversallanguageunderstanding-2-201202181619-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Keynote talk given at Pacific Asia Conference on Language, Information and Computation (PACLIC 34) on Pacific Asia Conference on Language, Information and Computation (PACLIC 34) on October 24, 2020. Title: Towards Universal Natural Language Understanding Abstract: Understanding the semantics of the natural language is a fundamental task in artificial intelligence. English semantic understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Compare and Comply. However, scaling existing products/services to support additional languages remain an open challenge. In this talk, we will discuss the open challenges in supporting universal natural language understanding. We will share our work in addressing these challenges in the past few years to provide the same unified semantic representation across languages. We will also showcase how such universal semantic understanding of natural languages can enable cross-lingual information extraction in concrete domains (e.g. insurance and compliance) and show promise towards seamless scaling existing NLP capabilities across languages with minimal efforts.
Towards Universal Language Understanding (2020 version) from Yunyao Li
]]> 138 0 https://cdn.slidesharecdn.com/ss_thumbnails/towardsuniversallanguageunderstanding-2-201202181619-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Towards Universal Semantic Understanding of Natural Languages /slideshow/towards-universal-language-understanding/218625110 towardsuniversallanguageunderstanding3-200111003040
Keynote talk at TextXD 2019(https://www.textxd.org) Abstract: Understanding the semantics of the natural language is a fundamental task in artificial intelligence. English semantic understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Compare and Comply. However, scaling existing products/services to support additional languages remain an open challenge. In this demo, we will present Polyglot, a multilingual semantic parser capable of semantically parsing sentences in 9 different languages from 4 different language groups into the same unified semantic representation. We will also showcase how such universal semantic understanding of natural languages can enable cross-lingual information extraction in concrete domains (e.g. insurance and compliance) and show promise towards seamless scaling existing NLP capabilities across languages with minimal efforts.]]>
Keynote talk at TextXD 2019(https://www.textxd.org) Abstract: Understanding the semantics of the natural language is a fundamental task in artificial intelligence. English semantic understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Compare and Comply. However, scaling existing products/services to support additional languages remain an open challenge. In this demo, we will present Polyglot, a multilingual semantic parser capable of semantically parsing sentences in 9 different languages from 4 different language groups into the same unified semantic representation. We will also showcase how such universal semantic understanding of natural languages can enable cross-lingual information extraction in concrete domains (e.g. insurance and compliance) and show promise towards seamless scaling existing NLP capabilities across languages with minimal efforts.]]> Sat, 11 Jan 2020 00:30:40 GMT /slideshow/towards-universal-language-understanding/218625110 YunyaoLi@slideshare.net(YunyaoLi) Towards Universal Semantic Understanding of Natural Languages YunyaoLi Keynote talk at TextXD 2019(https://www.textxd.org) Abstract: Understanding the semantics of the natural language is a fundamental task in artificial intelligence. English semantic understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Compare and Comply. However, scaling existing products/services to support additional languages remain an open challenge. In this demo, we will present Polyglot, a multilingual semantic parser capable of semantically parsing sentences in 9 different languages from 4 different language groups into the same unified semantic representation. We will also showcase how such universal semantic understanding of natural languages can enable cross-lingual information extraction in concrete domains (e.g. insurance and compliance) and show promise towards seamless scaling existing NLP capabilities across languages with minimal efforts. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/towardsuniversallanguageunderstanding3-200111003040-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Keynote talk at TextXD 2019(https://www.textxd.org) Abstract: Understanding the semantics of the natural language is a fundamental task in artificial intelligence. English semantic understanding has reached a mature state and successfully deployed in multiple IBM AI products and services, such as Watson Natural Language Understanding and Watson Compare and Comply. However, scaling existing products/services to support additional languages remain an open challenge. In this demo, we will present Polyglot, a multilingual semantic parser capable of semantically parsing sentences in 9 different languages from 4 different language groups into the same unified semantic representation. We will also showcase how such universal semantic understanding of natural languages can enable cross-lingual information extraction in concrete domains (e.g. insurance and compliance) and show promise towards seamless scaling existing NLP capabilities across languages with minimal efforts.
Towards Universal Semantic Understanding of Natural Languages from Yunyao Li
]]> 217 0 https://cdn.slidesharecdn.com/ss_thumbnails/towardsuniversallanguageunderstanding3-200111003040-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 An In-depth Analysis of the Effect of Text Normalization in Social Media /slideshow/an-indepth-analysis-of-the-effect-of-text-normalization-in-social-media/146770222 normalizationposternaacl2015final-190520172434
Poster corresponding to our NAACL'2015 paper "An In-depth Analysis of the Effect of Text Normalization in Social Media" Abstract: Recent years have seen increased interest in text normalization in social media, as the in-formal writing styles found in Twitter and other social media data often cause problems for NLP applications. Unfortunately, most current approaches narrowly regard the nor- malization task as a one size fits all" task of replacing non-standard words with their standard counterparts. In this work we build a taxonomy of normalization edits and present a study of normalization to examine its effect on three different downstream applications (de- pendency parsing, named entity recognition, and text-to-speech synthesis). The results sug- gest that how the normalization task should be viewed is highly dependent on the targeted application. The results also show that normalization must be thought of as more than word replacement in order to produce results comparable to those seen on clean text. Paper: https://www.aclweb.org/anthology/N15-1045]]>
Poster corresponding to our NAACL'2015 paper "An In-depth Analysis of the Effect of Text Normalization in Social Media" Abstract: Recent years have seen increased interest in text normalization in social media, as the in-formal writing styles found in Twitter and other social media data often cause problems for NLP applications. Unfortunately, most current approaches narrowly regard the nor- malization task as a one size fits all" task of replacing non-standard words with their standard counterparts. In this work we build a taxonomy of normalization edits and present a study of normalization to examine its effect on three different downstream applications (de- pendency parsing, named entity recognition, and text-to-speech synthesis). The results sug- gest that how the normalization task should be viewed is highly dependent on the targeted application. The results also show that normalization must be thought of as more than word replacement in order to produce results comparable to those seen on clean text. Paper: https://www.aclweb.org/anthology/N15-1045]]> Mon, 20 May 2019 17:24:33 GMT /slideshow/an-indepth-analysis-of-the-effect-of-text-normalization-in-social-media/146770222 YunyaoLi@slideshare.net(YunyaoLi) An In-depth Analysis of the Effect of Text Normalization in Social Media YunyaoLi Poster corresponding to our NAACL'2015 paper "An In-depth Analysis of the Effect of Text Normalization in Social Media" Abstract: Recent years have seen increased interest in text normalization in social media, as the in-formal writing styles found in Twitter and other social media data often cause problems for NLP applications. Unfortunately, most current approaches narrowly regard the nor- malization task as a one size fits all" task of replacing non-standard words with their standard counterparts. In this work we build a taxonomy of normalization edits and present a study of normalization to examine its effect on three different downstream applications (de- pendency parsing, named entity recognition, and text-to-speech synthesis). The results sug- gest that how the normalization task should be viewed is highly dependent on the targeted application. The results also show that normalization must be thought of as more than word replacement in order to produce results comparable to those seen on clean text. Paper: https://www.aclweb.org/anthology/N15-1045 <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/normalizationposternaacl2015final-190520172434-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Poster corresponding to our NAACL&#39;2015 paper &quot;An In-depth Analysis of the Effect of Text Normalization in Social Media&quot; Abstract: Recent years have seen increased interest in text normalization in social media, as the in-formal writing styles found in Twitter and other social media data often cause problems for NLP applications. Unfortunately, most current approaches narrowly regard the nor- malization task as a one size fits all&quot; task of replacing non-standard words with their standard counterparts. In this work we build a taxonomy of normalization edits and present a study of normalization to examine its effect on three different downstream applications (de- pendency parsing, named entity recognition, and text-to-speech synthesis). The results sug- gest that how the normalization task should be viewed is highly dependent on the targeted application. The results also show that normalization must be thought of as more than word replacement in order to produce results comparable to those seen on clean text. Paper: https://www.aclweb.org/anthology/N15-1045
An In-depth Analysis of the Effect of Text Normalization in Social Media from Yunyao Li
]]> 192 1 https://cdn.slidesharecdn.com/ss_thumbnails/normalizationposternaacl2015final-190520172434-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Exploiting Structure in Representation of Named Entities using Active Learning /slideshow/exploiting-structure-in-representation-of-named-entities-using-active-learning/110943517 lustre-coling-180822000831
際際滷s for our COLING'18 paper: http://aclweb.org/anthology/C18-1058 Fundamental to several knowledge-centric applications is the need to identify named entities from their textual mentions. However, entities lack a unique representation and their mentions can differ greatly. These variations arise in complex ways that cannot be captured using textual similarity metrics. However, entities have underlying structures, typically shared by entities of the same entity type, that can help reason over their name variations. Discovering, learning and manipulating these structures typically requires high manual effort in the form of large amounts of labeled training data and handwritten transformation programs. In this work, we propose an active-learning based framework that drastically reduces the labeled data required to learn the structures of entities. We show that programs for mapping entity mentions to their structures can be automatically generated using human-comprehensible labels. Our experiments show that our framework consistently outperforms both handwritten programs and supervised learning models. We also demonstrate the utility of our framework in relation extraction and entity resolution tasks.]]>
際際滷s for our COLING'18 paper: http://aclweb.org/anthology/C18-1058 Fundamental to several knowledge-centric applications is the need to identify named entities from their textual mentions. However, entities lack a unique representation and their mentions can differ greatly. These variations arise in complex ways that cannot be captured using textual similarity metrics. However, entities have underlying structures, typically shared by entities of the same entity type, that can help reason over their name variations. Discovering, learning and manipulating these structures typically requires high manual effort in the form of large amounts of labeled training data and handwritten transformation programs. In this work, we propose an active-learning based framework that drastically reduces the labeled data required to learn the structures of entities. We show that programs for mapping entity mentions to their structures can be automatically generated using human-comprehensible labels. Our experiments show that our framework consistently outperforms both handwritten programs and supervised learning models. We also demonstrate the utility of our framework in relation extraction and entity resolution tasks.]]> Wed, 22 Aug 2018 00:08:31 GMT /slideshow/exploiting-structure-in-representation-of-named-entities-using-active-learning/110943517 YunyaoLi@slideshare.net(YunyaoLi) Exploiting Structure in Representation of Named Entities using Active Learning YunyaoLi 際際滷s for our COLING'18 paper: http://aclweb.org/anthology/C18-1058 Fundamental to several knowledge-centric applications is the need to identify named entities from their textual mentions. However, entities lack a unique representation and their mentions can differ greatly. These variations arise in complex ways that cannot be captured using textual similarity metrics. However, entities have underlying structures, typically shared by entities of the same entity type, that can help reason over their name variations. Discovering, learning and manipulating these structures typically requires high manual effort in the form of large amounts of labeled training data and handwritten transformation programs. In this work, we propose an active-learning based framework that drastically reduces the labeled data required to learn the structures of entities. We show that programs for mapping entity mentions to their structures can be automatically generated using human-comprehensible labels. Our experiments show that our framework consistently outperforms both handwritten programs and supervised learning models. We also demonstrate the utility of our framework in relation extraction and entity resolution tasks. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/lustre-coling-180822000831-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> 際際滷s for our COLING&#39;18 paper: http://aclweb.org/anthology/C18-1058 Fundamental to several knowledge-centric applications is the need to identify named entities from their textual mentions. However, entities lack a unique representation and their mentions can differ greatly. These variations arise in complex ways that cannot be captured using textual similarity metrics. However, entities have underlying structures, typically shared by entities of the same entity type, that can help reason over their name variations. Discovering, learning and manipulating these structures typically requires high manual effort in the form of large amounts of labeled training data and handwritten transformation programs. In this work, we propose an active-learning based framework that drastically reduces the labeled data required to learn the structures of entities. We show that programs for mapping entity mentions to their structures can be automatically generated using human-comprehensible labels. Our experiments show that our framework consistently outperforms both handwritten programs and supervised learning models. We also demonstrate the utility of our framework in relation extraction and entity resolution tasks.
Exploiting Structure in Representation of Named Entities using Active Learning from Yunyao Li
]]> 174 4 https://cdn.slidesharecdn.com/ss_thumbnails/lustre-coling-180822000831-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 K-SRL: Instance-based Learning for Semantic Role Labeling /slideshow/k-srl-78770116/78770116 k-srl-170811170950
際際滷s for our COLING'16 paper http://aclweb.org/anthology/C/C16/C16-1058.pdf Abstract: Semantic role labeling (SRL) is the task of identifying and labeling predicate-argument structures in sentences with semantic frame and role labels. A known challenge in SRL is the large number of low-frequency exceptions in training data, which are highly context-specific and difficult to generalize. To overcome this challenge, we propose the use of instance-based learning that performs no explicit generalization, but rather extrapolates predictions from the most similar instances in the training data. We present a variant of k-nearest neighbors (kNN) classification with composite features to identify nearest neighbors for SRL. We show that high-quality predictions can be derived from a very small number of similar instances. In a comparative evaluation we experimentally demonstrate that our instance-based learning approach significantly outperforms current state-of-the-art systems on both in-domain and out-of-domain data, reaching F1-scores of 89,28% and 79.91% respectively]]>
際際滷s for our COLING'16 paper http://aclweb.org/anthology/C/C16/C16-1058.pdf Abstract: Semantic role labeling (SRL) is the task of identifying and labeling predicate-argument structures in sentences with semantic frame and role labels. A known challenge in SRL is the large number of low-frequency exceptions in training data, which are highly context-specific and difficult to generalize. To overcome this challenge, we propose the use of instance-based learning that performs no explicit generalization, but rather extrapolates predictions from the most similar instances in the training data. We present a variant of k-nearest neighbors (kNN) classification with composite features to identify nearest neighbors for SRL. We show that high-quality predictions can be derived from a very small number of similar instances. In a comparative evaluation we experimentally demonstrate that our instance-based learning approach significantly outperforms current state-of-the-art systems on both in-domain and out-of-domain data, reaching F1-scores of 89,28% and 79.91% respectively]]> Fri, 11 Aug 2017 17:09:50 GMT /slideshow/k-srl-78770116/78770116 YunyaoLi@slideshare.net(YunyaoLi) K-SRL: Instance-based Learning for Semantic Role Labeling YunyaoLi 際際滷s for our COLING'16 paper http://aclweb.org/anthology/C/C16/C16-1058.pdf Abstract: Semantic role labeling (SRL) is the task of identifying and labeling predicate-argument structures in sentences with semantic frame and role labels. A known challenge in SRL is the large number of low-frequency exceptions in training data, which are highly context-specific and difficult to generalize. To overcome this challenge, we propose the use of instance-based learning that performs no explicit generalization, but rather extrapolates predictions from the most similar instances in the training data. We present a variant of k-nearest neighbors (kNN) classification with composite features to identify nearest neighbors for SRL. We show that high-quality predictions can be derived from a very small number of similar instances. In a comparative evaluation we experimentally demonstrate that our instance-based learning approach significantly outperforms current state-of-the-art systems on both in-domain and out-of-domain data, reaching F1-scores of 89,28% and 79.91% respectively <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/k-srl-170811170950-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> 際際滷s for our COLING&#39;16 paper http://aclweb.org/anthology/C/C16/C16-1058.pdf Abstract: Semantic role labeling (SRL) is the task of identifying and labeling predicate-argument structures in sentences with semantic frame and role labels. A known challenge in SRL is the large number of low-frequency exceptions in training data, which are highly context-specific and difficult to generalize. To overcome this challenge, we propose the use of instance-based learning that performs no explicit generalization, but rather extrapolates predictions from the most similar instances in the training data. We present a variant of k-nearest neighbors (kNN) classification with composite features to identify nearest neighbors for SRL. We show that high-quality predictions can be derived from a very small number of similar instances. In a comparative evaluation we experimentally demonstrate that our instance-based learning approach significantly outperforms current state-of-the-art systems on both in-domain and out-of-domain data, reaching F1-scores of 89,28% and 79.91% respectively
K-SRL: Instance-based Learning for Semantic Role Labeling from Yunyao Li
]]> 341 5 https://cdn.slidesharecdn.com/ss_thumbnails/k-srl-170811170950-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Coling poster /slideshow/coling-poster/78770115 coling-poster-170811170950
Expert Curation]]>
Expert Curation]]> Fri, 11 Aug 2017 17:09:50 GMT /slideshow/coling-poster/78770115 YunyaoLi@slideshare.net(YunyaoLi) Coling poster YunyaoLi Expert Curation <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/coling-poster-170811170950-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Expert Curation
Coling poster from Yunyao Li
]]> 223 3 https://cdn.slidesharecdn.com/ss_thumbnails/coling-poster-170811170950-thumbnail.jpg?width=120&height=120&fit=bounds document Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Coling demo /slideshow/coling-demo/78770114 coling-demo-170811170949
Polyglot IE]]>
Polyglot IE]]> Fri, 11 Aug 2017 17:09:49 GMT /slideshow/coling-demo/78770114 YunyaoLi@slideshare.net(YunyaoLi) Coling demo YunyaoLi Polyglot IE <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/coling-demo-170811170949-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Polyglot IE
Coling demo from Yunyao Li
]]> 131 2 https://cdn.slidesharecdn.com/ss_thumbnails/coling-demo-170811170949-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Natural Language Data Management and Interfaces: Recent Development and Open Challenges /slideshow/natural-language-data-management-and-interfaces-recent-development-and-open-challenges/76924530 sigmod2017tutorial-170614050117
際際滷s deck for SIGMOD 2017 Tutorial. ABSTRACT: The volume of natural language text data has been rapidly increasing over the past two decades, due to factors such as the growth of the Web, the low cost associated to publishing and the progress on the digitization of printed texts. This growth combined with the proliferation of natural language systems for search and retrieving information provides tremendous opportunities for studying some of the areas where database systems and natural language processing systems overlap. This tutorial explores two more relevant areas of overlap to the database community: (1) managing natural language text data in a relational database, and (2) developing natural language interfaces to databases. The tutorial presents state-of-the-art methods, related systems, research opportunities and challenges covering both area.]]>
際際滷s deck for SIGMOD 2017 Tutorial. ABSTRACT: The volume of natural language text data has been rapidly increasing over the past two decades, due to factors such as the growth of the Web, the low cost associated to publishing and the progress on the digitization of printed texts. This growth combined with the proliferation of natural language systems for search and retrieving information provides tremendous opportunities for studying some of the areas where database systems and natural language processing systems overlap. This tutorial explores two more relevant areas of overlap to the database community: (1) managing natural language text data in a relational database, and (2) developing natural language interfaces to databases. The tutorial presents state-of-the-art methods, related systems, research opportunities and challenges covering both area.]]> Wed, 14 Jun 2017 05:01:17 GMT /slideshow/natural-language-data-management-and-interfaces-recent-development-and-open-challenges/76924530 YunyaoLi@slideshare.net(YunyaoLi) Natural Language Data Management and Interfaces: Recent Development and Open Challenges YunyaoLi 際際滷s deck for SIGMOD 2017 Tutorial. ABSTRACT: The volume of natural language text data has been rapidly increasing over the past two decades, due to factors such as the growth of the Web, the low cost associated to publishing and the progress on the digitization of printed texts. This growth combined with the proliferation of natural language systems for search and retrieving information provides tremendous opportunities for studying some of the areas where database systems and natural language processing systems overlap. This tutorial explores two more relevant areas of overlap to the database community: (1) managing natural language text data in a relational database, and (2) developing natural language interfaces to databases. The tutorial presents state-of-the-art methods, related systems, research opportunities and challenges covering both area. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/sigmod2017tutorial-170614050117-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> 際際滷s deck for SIGMOD 2017 Tutorial. ABSTRACT: The volume of natural language text data has been rapidly increasing over the past two decades, due to factors such as the growth of the Web, the low cost associated to publishing and the progress on the digitization of printed texts. This growth combined with the proliferation of natural language systems for search and retrieving information provides tremendous opportunities for studying some of the areas where database systems and natural language processing systems overlap. This tutorial explores two more relevant areas of overlap to the database community: (1) managing natural language text data in a relational database, and (2) developing natural language interfaces to databases. The tutorial presents state-of-the-art methods, related systems, research opportunities and challenges covering both area.
Natural Language Data Management and Interfaces: Recent Development and Open Challenges from Yunyao Li
]]> 903 5 https://cdn.slidesharecdn.com/ss_thumbnails/sigmod2017tutorial-170614050117-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 Polyglot: Multilingual Semantic Role Labeling with Unified Labels /slideshow/polyglot-multilingual-semantic-role-labeling-with-unified-labels/65846505 acl-poster-160909044433
Poster for our ACL paper "Polyglot: Multilingual Semantic Role Labeling with Unified Labels". Abstract: We present POLYGLOT, a semantic role labeling system capable of semantically parsing sentences in 9 different languages from 4 different language groups. A core differentiator is that this system predicts English Proposition Bank labels for all supported languages. This means that for instance a Japanese sentence will be tagged with the same labels as an English sentence with similar semantics would be. This is made possible by training the system with target language data that was automatically labeled with English PropBank labels using an annotation projection approach. We give an overview of our system, the automatically produced training data, and discuss possible applications and limitations of this work. We present a demonstrator that accepts sentences in English, German, French, Spanish, Japanese, Chinese, Arabic, Russian and Hindi and outputs a visualization of its shallow semantics. ]]>
Poster for our ACL paper "Polyglot: Multilingual Semantic Role Labeling with Unified Labels". Abstract: We present POLYGLOT, a semantic role labeling system capable of semantically parsing sentences in 9 different languages from 4 different language groups. A core differentiator is that this system predicts English Proposition Bank labels for all supported languages. This means that for instance a Japanese sentence will be tagged with the same labels as an English sentence with similar semantics would be. This is made possible by training the system with target language data that was automatically labeled with English PropBank labels using an annotation projection approach. We give an overview of our system, the automatically produced training data, and discuss possible applications and limitations of this work. We present a demonstrator that accepts sentences in English, German, French, Spanish, Japanese, Chinese, Arabic, Russian and Hindi and outputs a visualization of its shallow semantics. ]]> Fri, 09 Sep 2016 04:44:33 GMT /slideshow/polyglot-multilingual-semantic-role-labeling-with-unified-labels/65846505 YunyaoLi@slideshare.net(YunyaoLi) Polyglot: Multilingual Semantic Role Labeling with Unified Labels YunyaoLi Poster for our ACL paper "Polyglot: Multilingual Semantic Role Labeling with Unified Labels". Abstract: We present POLYGLOT, a semantic role labeling system capable of semantically parsing sentences in 9 different languages from 4 different language groups. A core differentiator is that this system predicts English Proposition Bank labels for all supported languages. This means that for instance a Japanese sentence will be tagged with the same labels as an English sentence with similar semantics would be. This is made possible by training the system with target language data that was automatically labeled with English PropBank labels using an annotation projection approach. We give an overview of our system, the automatically produced training data, and discuss possible applications and limitations of this work. We present a demonstrator that accepts sentences in English, German, French, Spanish, Japanese, Chinese, Arabic, Russian and Hindi and outputs a visualization of its shallow semantics. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/acl-poster-160909044433-thumbnail.jpg?width=120&amp;height=120&amp;fit=bounds" /><br> Poster for our ACL paper &quot;Polyglot: Multilingual Semantic Role Labeling with Unified Labels&quot;. Abstract: We present POLYGLOT, a semantic role labeling system capable of semantically parsing sentences in 9 different languages from 4 different language groups. A core differentiator is that this system predicts English Proposition Bank labels for all supported languages. This means that for instance a Japanese sentence will be tagged with the same labels as an English sentence with similar semantics would be. This is made possible by training the system with target language data that was automatically labeled with English PropBank labels using an annotation projection approach. We give an overview of our system, the automatically produced training data, and discuss possible applications and limitations of this work. We present a demonstrator that accepts sentences in English, German, French, Spanish, Japanese, Chinese, Arabic, Russian and Hindi and outputs a visualization of its shallow semantics.
Polyglot: Multilingual Semantic Role Labeling with Unified Labels from Yunyao Li
]]> 724 3 https://cdn.slidesharecdn.com/ss_thumbnails/acl-poster-160909044433-thumbnail.jpg?width=120&height=120&fit=bounds document 000000 http://activitystrea.ms/schema/1.0/post http://activitystrea.ms/schema/1.0/posted 0 https://cdn.slidesharecdn.com/profile-photo-YunyaoLi-48x48.jpg?cb=1740333768 Yunyao Li is the Director of Machine Learning at Adobe Experience Platform, where she leads the strategic initives to bring the power of Generative AI and Knowledge Graph to enterprise systems and transform the way companies approach audiences, journeys and personalization at scale. Previously, she was the Head of Machine Learning, Apple Knowledge Platform, Before joining Apple, she was a Distinguished Research Staff Member and Senior Research Manager at IBM Research - Almaden where she built and managed the Scalable Knowledge Intelligence department. She was also an IBM Master Inventor and a member of IBM Academy of Technology. yunyaoli.github.io https://cdn.slidesharecdn.com/ss_thumbnails/20meaning-representations-for-natural-languages-design-models-and-applications-240618162649-7129583c-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/meaning-representations-for-natural-languages-design-models-and-applications-pdf/269747868 Meaning Representation... https://cdn.slidesharecdn.com/ss_thumbnails/pandl-emnlp-december-2023-231206072657-815069fe-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/the-role-of-patterns-in-the-era-of-large-language-models/264351235 The Role of Patterns i... https://cdn.slidesharecdn.com/ss_thumbnails/hilda-keynote-june-18-2023-v1-230618213135-4a78f89d-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/building-growing-and-serving-large-knowledge-graphs-with-humanintheloop/258485723 Building, Growing and ...