�ݺ�ߣshows by User: YifanPeng

�ݺ�ߣshows by User: YifanPeng / http://www.slideshare.net/images/logo.gif �ݺ�ߣshows by User: YifanPeng / Thu, 15 Mar 2018 20:44:04 GMT �ݺ�ߣShare feed for �ݺ�ߣshows by User: YifanPeng NegBio: a high-performance tool for negation and uncertainty detection in radiology reports /slideshow/negbio-a-highperformance-tool-for-negation-and-uncertainty-detection-in-radiology-reports/90828254 amia2018-yifanpeng-180315204404
https://github.com/ncbi-nlp/NegBio]]>
https://github.com/ncbi-nlp/NegBio]]> Thu, 15 Mar 2018 20:44:04 GMT /slideshow/negbio-a-highperformance-tool-for-negation-and-uncertainty-detection-in-radiology-reports/90828254 YifanPeng@slideshare.net(YifanPeng) NegBio: a high-performance tool for negation and uncertainty detection in radiology reports YifanPeng https://github.com/ncbi-nlp/NegBio <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/amia2018-yifanpeng-180315204404-thumbnail.jpg?width=120&height=120&fit=bounds" /><br> https://github.com/ncbi-nlp/NegBio

NegBio: a high-performance tool for negation and uncertainty detection in radiology reports from Yifan Peng

]]> 283 3 https://cdn.slidesharecdn.com/ss_thumbnails/amia2018-yifanpeng-180315204404-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 Deep learning for extracting protein-protein interactions from biomedical literature /slideshow/deep-learning-for-extracting-proteinprotein-interactions-from-biomedical-literature/81686335 v2-yifanpeng-deeplearningforextractingprotein-proteininteractionsfrombiomedicalliterature-171107031725
State-of-the-art methods for protein-protein interaction (PPI) extraction are primarily feature-based or kernel-based by leveraging lexical and syntactic information. But how to incorporate such knowledge in the recent deep learning methods remains an open question. In this paper, we propose a multichannel dependency-based convolutional neural network model (McDepCNN). It applies one channel to the embedding vector of each word in the sentence, and another channel to the embedding vector of the head of the corresponding word. Therefore, the model can use richer information obtained from different channels. Experiments on two public benchmarking datasets, AIMed and BioInfer, demonstrate that McDepCNN compares favorably to the state-of-the-art rich-feature and single-kernel based methods. In addition, McDepCNN achieves 24.4% relative improvement in F1-score over the state-of-the-art methods on cross-corpus evaluation and 12% improvement in F1-score over kernel-based methods on "difficult" instances. These results suggest that McDepCNN generalizes more easily over different corpora, and is capable of capturing long distance features in the sentences.]]>
State-of-the-art methods for protein-protein interaction (PPI) extraction are primarily feature-based or kernel-based by leveraging lexical and syntactic information. But how to incorporate such knowledge in the recent deep learning methods remains an open question. In this paper, we propose a multichannel dependency-based convolutional neural network model (McDepCNN). It applies one channel to the embedding vector of each word in the sentence, and another channel to the embedding vector of the head of the corresponding word. Therefore, the model can use richer information obtained from different channels. Experiments on two public benchmarking datasets, AIMed and BioInfer, demonstrate that McDepCNN compares favorably to the state-of-the-art rich-feature and single-kernel based methods. In addition, McDepCNN achieves 24.4% relative improvement in F1-score over the state-of-the-art methods on cross-corpus evaluation and 12% improvement in F1-score over kernel-based methods on "difficult" instances. These results suggest that McDepCNN generalizes more easily over different corpora, and is capable of capturing long distance features in the sentences.]]> Tue, 07 Nov 2017 03:17:25 GMT /slideshow/deep-learning-for-extracting-proteinprotein-interactions-from-biomedical-literature/81686335 YifanPeng@slideshare.net(YifanPeng) Deep learning for extracting protein-protein interactions from biomedical literature YifanPeng State-of-the-art methods for protein-protein interaction (PPI) extraction are primarily feature-based or kernel-based by leveraging lexical and syntactic information. But how to incorporate such knowledge in the recent deep learning methods remains an open question. In this paper, we propose a multichannel dependency-based convolutional neural network model (McDepCNN). It applies one channel to the embedding vector of each word in the sentence, and another channel to the embedding vector of the head of the corresponding word. Therefore, the model can use richer information obtained from different channels. Experiments on two public benchmarking datasets, AIMed and BioInfer, demonstrate that McDepCNN compares favorably to the state-of-the-art rich-feature and single-kernel based methods. In addition, McDepCNN achieves 24.4% relative improvement in F1-score over the state-of-the-art methods on cross-corpus evaluation and 12% improvement in F1-score over kernel-based methods on "difficult" instances. These results suggest that McDepCNN generalizes more easily over different corpora, and is capable of capturing long distance features in the sentences. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/v2-yifanpeng-deeplearningforextractingprotein-proteininteractionsfrombiomedicalliterature-171107031725-thumbnail.jpg?width=120&height=120&fit=bounds" /><br> State-of-the-art methods for protein-protein interaction (PPI) extraction are primarily feature-based or kernel-based by leveraging lexical and syntactic information. But how to incorporate such knowledge in the recent deep learning methods remains an open question. In this paper, we propose a multichannel dependency-based convolutional neural network model (McDepCNN). It applies one channel to the embedding vector of each word in the sentence, and another channel to the embedding vector of the head of the corresponding word. Therefore, the model can use richer information obtained from different channels. Experiments on two public benchmarking datasets, AIMed and BioInfer, demonstrate that McDepCNN compares favorably to the state-of-the-art rich-feature and single-kernel based methods. In addition, McDepCNN achieves 24.4% relative improvement in F1-score over the state-of-the-art methods on cross-corpus evaluation and 12% improvement in F1-score over kernel-based methods on "difficult" instances. These results suggest that McDepCNN generalizes more easily over different corpora, and is capable of capturing long distance features in the sentences.

Deep learning for extracting protein-protein interactions from biomedical literature from Yifan Peng

]]> 303 2 https://cdn.slidesharecdn.com/ss_thumbnails/v2-yifanpeng-deeplearningforextractingprotein-proteininteractionsfrombiomedicalliterature-171107031725-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Chemical-protein relation extraction with ensembles of SVM, CNN, and RNN models /slideshow/chemicalprotein-relation-extraction-with-ensembles-of-svm-cnn-and-rnn-models-81686260/81686260 v3-biocreative6-171107031440
Text mining the relations between chemicals and proteins is an increasingly important task. The CHEMPROT track at BioCreative VI aims to promote the development and evaluation of systems that can automatically detect the chemical-protein relations in running text (PubMed abstracts). This manuscript describes our submission, which is an ensemble of three systems, including a Support Vector Machine, a Convolutional Neural Network, and a Recurrent Neural Network. Their output is combined using a decision based on majority voting or stacking. Our CHEMPROT system obtained 0.7266 in precision and 0.5735 in recall for an f-score of 0.6410, demonstrating the effectiveness of machine learning-based approaches for automatic relation extraction from biomedical literature. Our submission achieved the highest performance in the task during the 2017 challenge.]]>
Text mining the relations between chemicals and proteins is an increasingly important task. The CHEMPROT track at BioCreative VI aims to promote the development and evaluation of systems that can automatically detect the chemical-protein relations in running text (PubMed abstracts). This manuscript describes our submission, which is an ensemble of three systems, including a Support Vector Machine, a Convolutional Neural Network, and a Recurrent Neural Network. Their output is combined using a decision based on majority voting or stacking. Our CHEMPROT system obtained 0.7266 in precision and 0.5735 in recall for an f-score of 0.6410, demonstrating the effectiveness of machine learning-based approaches for automatic relation extraction from biomedical literature. Our submission achieved the highest performance in the task during the 2017 challenge.]]> Tue, 07 Nov 2017 03:14:40 GMT /slideshow/chemicalprotein-relation-extraction-with-ensembles-of-svm-cnn-and-rnn-models-81686260/81686260 YifanPeng@slideshare.net(YifanPeng) Chemical-protein relation extraction with ensembles of SVM, CNN, and RNN models YifanPeng Text mining the relations between chemicals and proteins is an increasingly important task. The CHEMPROT track at BioCreative VI aims to promote the development and evaluation of systems that can automatically detect the chemical-protein relations in running text (PubMed abstracts). This manuscript describes our submission, which is an ensemble of three systems, including a Support Vector Machine, a Convolutional Neural Network, and a Recurrent Neural Network. Their output is combined using a decision based on majority voting or stacking. Our CHEMPROT system obtained 0.7266 in precision and 0.5735 in recall for an f-score of 0.6410, demonstrating the effectiveness of machine learning-based approaches for automatic relation extraction from biomedical literature. Our submission achieved the highest performance in the task during the 2017 challenge. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/v3-biocreative6-171107031440-thumbnail.jpg?width=120&height=120&fit=bounds" /><br> Text mining the relations between chemicals and proteins is an increasingly important task. The CHEMPROT track at BioCreative VI aims to promote the development and evaluation of systems that can automatically detect the chemical-protein relations in running text (PubMed abstracts). This manuscript describes our submission, which is an ensemble of three systems, including a Support Vector Machine, a Convolutional Neural Network, and a Recurrent Neural Network. Their output is combined using a decision based on majority voting or stacking. Our CHEMPROT system obtained 0.7266 in precision and 0.5735 in recall for an f-score of 0.6410, demonstrating the effectiveness of machine learning-based approaches for automatic relation extraction from biomedical literature. Our submission achieved the highest performance in the task during the 2017 challenge.

Chemical-protein relation extraction with ensembles of SVM, CNN, and RNN models from Yifan Peng

]]> 248 3 https://cdn.slidesharecdn.com/ss_thumbnails/v3-biocreative6-171107031440-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Text Mining Radiology Reports for Deep Learning Radiology Images /slideshow/text-mining-radiology-reports-for-deep-learning-radiology-images/81686101 final-amia2017-171107030933
Although hospitals have accumulated a tremendous number of radiology images and reports, using them to build data-hungry deep-learning models remains challenging. This is partially because manually annotating a dataset in large scale is costly. Here, we propose using text mining to automatically generate a weakly-labeled dataset for detecting thoracic diseases in radiology images. This work represents the first attempt to unify text mining with computer vision for medical imaging analysis in the era of deep learning.]]>
Although hospitals have accumulated a tremendous number of radiology images and reports, using them to build data-hungry deep-learning models remains challenging. This is partially because manually annotating a dataset in large scale is costly. Here, we propose using text mining to automatically generate a weakly-labeled dataset for detecting thoracic diseases in radiology images. This work represents the first attempt to unify text mining with computer vision for medical imaging analysis in the era of deep learning.]]> Tue, 07 Nov 2017 03:09:33 GMT /slideshow/text-mining-radiology-reports-for-deep-learning-radiology-images/81686101 YifanPeng@slideshare.net(YifanPeng) Text Mining Radiology Reports for Deep Learning Radiology Images YifanPeng Although hospitals have accumulated a tremendous number of radiology images and reports, using them to build data-hungry deep-learning models remains challenging. This is partially because manually annotating a dataset in large scale is costly. Here, we propose using text mining to automatically generate a weakly-labeled dataset for detecting thoracic diseases in radiology images. This work represents the first attempt to unify text mining with computer vision for medical imaging analysis in the era of deep learning. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/final-amia2017-171107030933-thumbnail.jpg?width=120&height=120&fit=bounds" /><br> Although hospitals have accumulated a tremendous number of radiology images and reports, using them to build data-hungry deep-learning models remains challenging. This is partially because manually annotating a dataset in large scale is costly. Here, we propose using text mining to automatically generate a weakly-labeled dataset for detecting thoracic diseases in radiology images. This work represents the first attempt to unify text mining with computer vision for medical imaging analysis in the era of deep learning.

Text Mining Radiology Reports for Deep Learning Radiology Images from Yifan Peng

]]> 360 3 https://cdn.slidesharecdn.com/ss_thumbnails/final-amia2017-171107030933-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 https://public.slidesharecdn.com/v2/images/profile-picture.png Relation extraction, Bioinformatics, Text mining, Natural language processing Specialties: natural language processing, bioinformatics http://www.eecis.udel.edu/~ypeng https://cdn.slidesharecdn.com/ss_thumbnails/amia2018-yifanpeng-180315204404-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/negbio-a-highperformance-tool-for-negation-and-uncertainty-detection-in-radiology-reports/90828254 NegBio: a high-perform... https://cdn.slidesharecdn.com/ss_thumbnails/v2-yifanpeng-deeplearningforextractingprotein-proteininteractionsfrombiomedicalliterature-171107031725-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/deep-learning-for-extracting-proteinprotein-interactions-from-biomedical-literature/81686335 Deep learning for extr... https://cdn.slidesharecdn.com/ss_thumbnails/v3-biocreative6-171107031440-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/chemicalprotein-relation-extraction-with-ensembles-of-svm-cnn-and-rnn-models-81686260/81686260 Chemical-protein relat...