Variational Template Machine for Data-to-Text Generationharmonylab
?
公開URL:https://openreview.net/forum?id=HkejNgBtPB
出典:Rong Ye, Wenxian Shi, Hao Zhou, Zhongyu Wei, Lei Li : Variational Template Machine for Data-to-Text Generation, 8th International Conference on Learning Representations(ICLR2020), Addis Ababa, Ethiopia (2020)
概要:Table形式の構造化データから文章を生成するタスク(Data-to-Text)において、Variational Auto Encoder(VAE)ベースの手法Variational Template Machine(VTM)を提案する論文です。Encoder-Decoderモデルを用いた既存のアプローチでは、生成文の多様性に欠けるという課題があります。本論文では多様な文章を生成するためにはテンプレートが重要であるという主張に基づき、テンプレートを学習可能なVAEベースの手法を提案します。提案手法では潜在変数の空間をテンプレート空間とコンテンツ空間に明示的に分離することによって、正確で多様な文生成が可能となります。また、table-textのペアデータだけではなくtableデータのないraw textデータを利用した半教師あり学習を行います。
Modeling missing data in distant supervision for information extraction (Ritt...Naoaki Okazaki
?
This document summarizes the MultiR model for distant supervision relation extraction. MultiR introduces latent variables to indicate the relation expressed by each sentence and handles missing data by relaxing hard constraints from previous models. It allows an entity pair to have multiple relations and incorporates the tendency that knowledge bases include popular entities and relations. The model is trained using an algorithm similar to perceptron and inference involves finding the highest weight assignment of relations consistent with the knowledge base.
Learning to automatically solve algebra word problemsNaoaki Okazaki
?
Nate Kushman, Yoav Artzi, Luke Zettlemoyer, and Regina Barzilay.
ACL-2014, pages 271–281.
(presented by Naoaki Okazaki at the paper reading organized by Preferred Infrastructure)
IoT Devices Compliant with JC-STAR Using Linux as a Container OSTomohiro Saneyoshi
?
Security requirements for IoT devices are becoming more defined, as seen with the EU Cyber Resilience Act and Japan’s JC-STAR.
It's common for IoT devices to run Linux as their operating system. However, adopting general-purpose Linux distributions like Ubuntu or Debian, or Yocto-based Linux, presents certain difficulties. This article outlines those difficulties.
It also, it highlights the security benefits of using a Linux-based container OS and explains how to adopt it with JC-STAR, using the "Armadillo Base OS" as an example.
Feb.25.2025@JAWS-UG IoT
9. 議論マイニングのボトルネックは世界知識
(Saint-Dizier 2016, Hanawa+ 2017, Moens 2018)
? 世界知識が必要となる例 (cellphoneの知識が要る)
? is-a(cellphone, technology)
? used-for(cellphone, communication)
? 議論マイニングにおいて世界知識は必須
? 議論単位間の関係予測において,78%の議論で世界知識が
必要 (Saint-Dizier 2016)
? 賛否分類タスクにおいて,33.9%の発言に対して世界知識
が必要 (Hanawa+ 2017)
2018-06-05 自然言语処理による议论マイニング 9
Technology negatively influences how people communicate.
Some people use their cellphone constantly and do not even notice their environment.
2番目の文は1番目の
文の論拠
11. 主張間の暗黙のギャップ (Boltuzic+ 2016)
2018-06-05 自然言语処理による议论マイニング 11
Legalized marijuana can be controlled and regulated by the government.
If something is not taxed, criminals sell it.
Things that are taxed are controlled and regulated by the government.
Criminals should be stopped from selling things.
Marijuana is not taxed, and those who sell it are usually criminals of some sort.
上の主張と同じ立場であることは人間には分かるが,コンピュータには難しい
ギャップを埋める暗黙の前提
13. Argument reasoning comprehension
(Habernal+ 2018) (ギャップを埋めるタスクを選択式とした)
2018-06-05 自然言语処理による议论マイニング 13
Russia cannot be a partner.
Russia has the same
objectives of the US.
Russia has the opposite
objectives of the US.
Cooperating with Russia on terrorism ignores Russia’s overall objectives.
理由
(reason)
主張
(claim)
根拠
(warrant)
Russia can be a partner.
? 理由と主張に対して,2つの根拠が与えられる
? どちらの根拠が理由と主張を繋ぐのにふさわしいか選ぶ
? もう一方の根拠は,理由から「反対の主張」を導くとして用意された
? この研究では反対の根拠(AW: alternative warrant)と呼んでいる
R
WAW
C
28. 参考文献
? Filip Boltuzic, Jan ?najder. 2016. Fill the gap! Analyzing implicit premises between claims from
online debates. Proc. of ArgMining, pp. 124-133.
? Marie-Francine Moens. 2018. Argumentation mining: How can a machine acquire common sense
and world knowledge? Argument & Computation, 9:1–14.
? Iryna Gurevych, Chris Reed, Noam Slonim, Benno Stein. NLP approaches to computational
argumentation. ACL 2016 tutorial.
? Ivan Habernal, Henning Wachsmuth, Iryna Gurevych, Benno Stein. 2018. The argument reasoning
comprehension task: Identification and reconstruction of implicit warrants. Proc. of NAACL-HLT, pp.
1930-1940.
? Kazuaki Hanawa, Akira Sasaki, Naoaki Okazaki, Kentaro Inui. 2017. A crowdsourcing approach for
annotating causal relation instances inWikipedia. Proc. of PACLIC.
? Kazi Saidul Hasan, Vincent Ng. 2014. Why are you taking this stance? Identifying and classifying
reasons in ideological debates. Proc. of EMNLP, pages 751-762.
? Saif Mohammad, Svetlana Kiritchenko, Parinaz Sobhani, Xiaodan Zhu, Colin Cherry. 2016. Semeval-
2016 task 6: Detecting stance in tweets. Proc. of SemEval, pp. 31-41.
? Patrick Saint-Dizier. 2016. Challenges of argument mining: Generating an argument synthesis
based on the qualia structure. Proc. of INLG, pp. 79-83.
? Akira Sasaki, Kazuaki Hanawa, Naoaki Okazaki, Kentaro Inui. 2017. Other topics you may also
agree or disagree: Modeling inter-topic preferences using tweets and matrix factorization. Proc. of
ACL, pp. 398-408.
? Akira Sasaki, Kazuaki Hanawa, Naoaki Okazaki, Kentaro Inui. 2017. Predicting stances from social
media posts using factorization machines. Proc. of Coling, (to appear).
? Christian Stab, Iryna Gurevych. 2017. Parsing argumentation structures in persuasive essays.
Computational Linguistics, 43(3):619-659.
2018-06-05 自然言语処理による议论マイニング 28