2024/03/12 更新

写真a

タケダ コウイチ
武田 浩一
TAKEDA Koichi
所属
大学院情報学研究科 附属価値創造研究センター 教授
大学院担当
大学院情報学研究科
学部担当
情報学部 コンピュータ科学科
職名
教授

学位 1

  1. 博士(情報学) ( 2010年3月   京都大学 ) 

研究キーワード 3

  1. 質問応答

  2. テキストマイニング

  3. 自然言語処理

研究分野 1

  1. 情報通信 / 知能情報学

現在の研究課題とSDGs 1

  1. 説明可能な人工知能のための自然言語処理

経歴 4

  1. 名古屋大学   情報学研究科 附属価値創造研究センター   教授

    2017年4月 - 現在

  2. 日本アイ・ビー・エム株式会社   東京基礎研究所   技術理事

    2011年5月 - 2017年3月

      詳細を見る

    国名:日本国

  3. 日本アイ・ビー・エム株式会社   東京基礎研究所   主席研究員

    2006年1月 - 2011年4月

      詳細を見る

    国名:日本国

  4. 日本アイ・ビー・エム株式会社   東京基礎研究所   研究員

    1983年4月 - 2005年12月

      詳細を見る

    国名:日本国

学歴 3

  1. 京都大学   情報学研究科   知能情報学専攻

    2007年4月 - 2010年3月

      詳細を見る

    国名: 日本国

  2. 京都大学   工学研究科   情報工学専攻

    1981年4月 - 1983年3月

      詳細を見る

    国名: 日本国

  3. 京都大学   工学部   情報工学科

    1977年4月 - 1981年3月

      詳細を見る

    国名: 日本国

所属学協会 4

  1. Association for Computing Machinery (ACM)

  2. 言語処理学会

  3. 電子情報通信学会

  4. 情報処理学会

受賞 5

  1. 楽天テクノロジーアワード金賞

    2016年10月   楽天株式会社  

     詳細を見る

    受賞国:日本国

  2. Feigenbaum Prize

    2013年7月  

     詳細を見る

    受賞区分:国際学会・会議・シンポジウム等の賞  受賞国:アメリカ合衆国

  3. 喜安記念業績賞

    2013年6月   情報処理学会  

     詳細を見る

    受賞区分:国内学会・会議・シンポジウム等の賞  受賞国:日本国

  4. 文部科学大臣表彰(科学技術賞 開発部門)

    2012年4月   文部科学省  

     詳細を見る

    受賞国:日本国

  5. 学術奨励賞

    1985年3月   情報処理学会  

     詳細を見る

    受賞区分:国内学会・会議・シンポジウム等の賞  受賞国:日本国

 

論文 35

  1. Inference Discrepancy Based Curriculum Learning for Neural Machine Translation

    Zhou, L; Sasano, R; Takeda, K

    IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS   E107.D 巻 ( 1 ) 頁: 135 - 143   2024年1月

     詳細を見る

    記述言語:英語   出版者・発行元:一般社団法人 電子情報通信学会  

    In practice, even a well-trained neural machine translation (NMT) model can still make biased inferences on the training set due to distribution shifts. For the human learning process, if we can not reproduce something correctly after learning it multiple times, we consider it to be more difficult. Likewise, a training example causing a large discrepancy between inference and reference implies higher learning difficulty for the MT model. Therefore, we propose to adopt the inference discrepancy of each training example as the difficulty criterion, and according to which rank training examples from easy to hard. In this way, a trained model can guide the curriculum learning process of an initial model identical to itself. We put forward an analogy to this training scheme as guiding the learning process of a curriculum NMT model by a pretrained vanilla model. In this paper, we assess the effectiveness of the proposed training scheme and take an insight into the influence of translation direction, evaluation metrics and different curriculum schedules. Experimental results on translation benchmarks WMT14 English → German, WMT17 Chinese → English and Multitarget TED Talks Task (MTTT) English ↔ German, English ↔ Chinese, English ↔ Russian demonstrate that our proposed method consistently improves the translation performance against the advanced Transformer baseline.

    DOI: 10.1587/transinf.2023edp7048

    Web of Science

    Scopus

    CiNii Research

  2. 深層距離学習を用いた意味フレーム構築におけるフレーム要素知識の獲得

    山田 康輔, 笹野 遼平, 武田 浩一

    人工知能学会全国大会論文集   JSAI2023 巻 ( 0 ) 頁: 3A1GS601 - 3A1GS601   2023年

     詳細を見る

    記述言語:日本語   出版者・発行元:一般社団法人 人工知能学会  

    <p>意味フレームを自動構築するためには、フレーム喚起語をその語が喚起するフレームごとにまとめること、およびその項となる語句をフレーム要素ごとにまとめる必要がある。本研究では、後者の項クラスタリングに焦点を当て、深層距離学習を用いた手法を提案する。提案手法では、フレーム情報が付与された一部のデータを用いて、深層距離学習に基づき文脈化単語埋め込みモデルをfine-tuningし、そのモデルから得られる埋め込み表現を利用して、動詞の項となる語句をクラスタリングしている。FrameNetを用いた実験を通し、提案手法が既存手法よりもかなり高い性能を達成することを示す。</p>

    DOI: 10.11517/pjsai.jsai2023.0_3a1gs601

    CiNii Research

  3. 言語モデルT5を用いたTwitterからのスポーツ速報生成

    大鹿 雅史, 山田 康輔, 笹野 遼平, 武田 浩一

    人工知能学会全国大会論文集   JSAI2023 巻 ( 0 ) 頁: 4Xin120 - 4Xin120   2023年

     詳細を見る

    記述言語:日本語   出版者・発行元:一般社団法人 人工知能学会  

    <p>テレビやインターネットで中継されるスポーツの試合中には、試合に関する多くのツイートがTwitterに投稿され、これらのツイートを読むことで大まかな試合経過の把握が可能である。しかし、ツイートの内容は試合についての情報を多く含むものから個人の感情を表すものまで多岐に渡り、これらのツイートから瞬時に試合経過を得ることは容易ではない。そこで、本稿では特にサッカーの試合に着目し、試合経過を瞬時に把握できるようにツイートからスポーツ実況を生成するシステムの構築に取り組む。具体的には、大規模言語モデルT5に基づく速報生成モデルに、速報生成数の制御および冗長性軽減を行う機構を組み込んだモデルを提案し、実験を通じてその有効性を示す。</p>

    DOI: 10.11517/pjsai.jsai2023.0_4xin120

    CiNii Research

  4. 定義文を用いた文埋め込み構成法

    塚越 駿, 笹野 遼平, 武田 浩一

    自然言語処理   30 巻 ( 1 ) 頁: 125 - 155   2023年

     詳細を見る

    記述言語:日本語   出版者・発行元:一般社団法人 言語処理学会  

    <p>自然言語文をベクトルとして表現する文埋め込みは,深層学習を用いた自然言語処理の基礎技術として盛んに研究されており,特に自然言語推論 (Natural Language Inference; NLI) タスクに基づく文埋め込み手法が成功を収めている.しかし,これらの手法は大規模な NLI データセットを必要とすることから,そのような NLI データが整備された言語以外については高品質な文埋め込みの構築が期待できないという問題がある.本研究ではこの問題を解決するため,NLI データと比べて多くの言語において整備が行われている言語資源である辞書に着目し,辞書の定義文を用いた文埋め込み手法を提案する.また,標準的なベンチマークを用いた評価実験を通し,提案手法は既存の NLI タスクに基づく文埋め込み手法と同等の性能を実現すること,評価タスクの性質や評価データの抽出方法により性能に差異が見られること,これら2手法を統合することでより高い性能を実現できることを示す.</p>

    DOI: 10.5715/jnlp.30.125

    CiNii Research

  5. Realistic Citation Count Prediction Task for Newly Published Papers

    Hirako J., Sasano R., Takeda K.

    EACL 2023 - 17th Conference of the European Chapter of the Association for Computational Linguistics, Findings of EACL 2023     頁: 1101 - 1111   2023年

     詳細を見る

    出版者・発行元:EACL 2023 - 17th Conference of the European Chapter of the Association for Computational Linguistics, Findings of EACL 2023  

    Citation count prediction is the task of predicting the future citation counts of academic papers, which is particularly useful for estimating the future impacts of an ever-growing number of academic papers. Although there have been many studies on citation count prediction, they are not applicable to predicting the citation counts of newly published papers, because they assume the availability of future citation counts for papers that have not had enough time pass since publication. In this paper, we first identify problems in the settings of existing studies and introduce a realistic citation count prediction task that strictly uses information available at the time of a target paper’s publication. For realistic citation count prediction, we then propose two methods to leverage the citation counts of papers shortly after publication. Through experiments using papers collected from arXiv and bioRxiv, we demonstrate that our methods considerably improve the performance of citation count prediction for newly published papers in a realistic setting.

    Scopus

  6. Semantic Frame Induction with Deep Metric Learning

    Yamada K., Sasano R., Takeda K.

    EACL 2023 - 17th Conference of the European Chapter of the Association for Computational Linguistics, Proceedings of the Conference     頁: 1825 - 1837   2023年

     詳細を見る

    出版者・発行元:EACL 2023 - 17th Conference of the European Chapter of the Association for Computational Linguistics, Proceedings of the Conference  

    Recent studies have demonstrated the usefulness of contextualized word embeddings in unsupervised semantic frame induction. However, they have also revealed that generic contextualized embeddings are not always consistent with human intuitions about semantic frames, which causes unsatisfactory performance for frame induction based on contextualized embeddings. In this paper, we address supervised semantic frame induction, which assumes the existence of frame-annotated data for a subset of predicates in a corpus and aims to build a frame induction model that leverages the annotated data. We propose a model that uses deep metric learning to fine-tune a contextualized embedding model, and we apply the fine-tuned contextualized embeddings to perform semantic frame induction. Our experiments on FrameNet show that fine-tuning with deep metric learning considerably improves the clustering evaluation scores, namely, the B-CUBED F-SCORE and PURITY F-SCORE, by about 8 points or more. We also demonstrate that our approach is effective even when the number of training instances is small.

    Scopus

  7. 深層距離学習を用いた動詞の意味フレーム推定

    山田 康輔, 笹野 遼平, 武田 浩一

    自然言語処理   30 巻 ( 4 ) 頁: 1130 - 1150   2023年

     詳細を見る

    記述言語:日本語   出版者・発行元:一般社団法人 言語処理学会  

    <p>意味フレーム推定において,文脈化単語埋め込みを用いる手法が高い性能を達成することが報告されている.しかし,汎用的な埋め込み空間は,意味的に類似したフレームの事例が近くに位置しているという人間の直観と必ずしも一致しているわけではないため,事前学習のみに基づく文脈化単語埋め込みを用いる手法の性能には限界がある.そこで,本研究では,意味フレーム推定をコーパス内の一部の動詞についてのラベル付きデータの存在を仮定した教師ありタスクとして取り組み,深層距離学習に基づき文脈化単語埋め込みモデルを fine-tuning することで高精度な意味フレーム推定を実現する手法を提案する.FrameNet を用いた実験を通し,深層距離学習を適用することで 8 ポイント以上スコアが向上することを示す.さらに,教師データが極めて少量である場合でも,提案手法が有効であることを示す.</p>

    DOI: 10.5715/jnlp.30.1130

    CiNii Research

  8. Acquiring Frame Element Knowledge with Deep Metric Learning for Semantic Frame Induction

    Yamada K., Sasano R., Takeda K.

    Proceedings of the Annual Meeting of the Association for Computational Linguistics     頁: 9356 - 9364   2023年

     詳細を見る

    出版者・発行元:Proceedings of the Annual Meeting of the Association for Computational Linguistics  

    The semantic frame induction tasks are defined as a clustering of words into the frames that they evoke, and a clustering of their arguments according to the frame element roles that they should fill. In this paper, we address the latter task of argument clustering, which aims to acquire frame element knowledge, and propose a method that applies deep metric learning. In this method, a pre-trained language model is fine-tuned to be suitable for distinguishing frame element roles through the use of frame-annotated data, and argument clustering is performed with embeddings obtained from the fine-tuned model. Experimental results on FrameNet demonstrate that our method achieves substantially better performance than existing methods.

    Scopus

  9. Building a Buzzer-Quiz Answering System

    Sugiura N., Yamada K., Sasano R., Takeda K., Toyama K.

    Proceedings of the Annual Meeting of the Association for Computational Linguistics   4 巻   頁: 194 - 199   2023年

     詳細を見る

    出版者・発行元:Proceedings of the Annual Meeting of the Association for Computational Linguistics  

    A buzzer quiz is a genre of quiz in which multiple players simultaneously listen to a quiz being read aloud and respond it by buzzing in as soon as they can predict the answer. Because incorrect answers often result in penalties, a buzzer-quiz answering system must not only predict the answer from only part of a question but also estimate the predicted answer's accuracy. In this paper, we introduce two types of buzzer-quiz answering systems: (1) a system that directly generates an answer from part of a question by using an autoregressive language model; and (2) a system that first reconstructs the entire question by using an autoregressive language model and then determines the answer according to the reconstructed question. We then propose a method to estimate the accuracy of the answers for each system by using the internal scores of each model.

    Scopus

  10. Transformer-based Live Update Generation for Soccer Matches from Microblog Posts

    Oshika M., Yamada K., Sasano R., Takeda K.

    EMNLP 2023 - 2023 Conference on Empirical Methods in Natural Language Processing, Proceedings     頁: 10100 - 10106   2023年

     詳細を見る

    出版者・発行元:EMNLP 2023 - 2023 Conference on Empirical Methods in Natural Language Processing, Proceedings  

    It has been known to be difficult to generate adequate sports updates from a sequence of vast amounts of diverse live tweets, although the live sports viewing experience with tweets is gaining the popularity. In this paper, we focus on soccer matches and work on building a system to generate live updates for soccer matches from tweets so that users can instantly grasp a match's progress and enjoy the excitement of the match from raw tweets. Our proposed system is based on a large pre-trained language model and incorporates a mechanism to control the number of updates and a mechanism to reduce the redundancy of duplicate and similar updates.

    Scopus

  11. マスクされた単語埋め込みと 2 段階クラスタリングを用いた動詞の意味フレーム推定

    山田 康輔, 笹野 遼平, 武田 浩一

    自然言語処理   29 巻 ( 2 ) 頁: 395 - 415   2022年

     詳細を見る

    記述言語:日本語   出版者・発行元:一般社団法人 言語処理学会  

    <p> 近年,動詞の意味フレーム推定タスクでは,推定対象の動詞の文脈化単語埋め込みに基づき,動詞全体で一度にクラスタリングを行う手法がいくつか提案されている.しかし,このような手法には大きく 2 つの欠点が存在する.1 つは動詞の表層的な情報を過度に考慮するため,意味の似た異なる動詞の用例をまとめづらいこと,もう 1 つは同じ動詞の用例がその動詞自身が持つ意味の異なり数以上のクラスタに分割されることである.本論文では,これらの欠点を克服するために,マスクされた単語埋め込みと 2 段階クラスタリングを用いた動詞の意味フレーム推定手法を提案する.FrameNet を用いた実験を通し,マスクされた単語埋め込みを活用することが動詞の表層的な情報に強く依存したクラスタの構築を抑制し,また,2 段階のクラスタリングを行うことで各動詞の用例が属するクラスタの異なり数を適正化できることを示す.</p>

    DOI: 10.5715/jnlp.29.395

    CiNii Research

  12. Comparison and Combination of Sentence Embeddings Derived from Different Supervision Signals

    Tsukagoshi H., Sasano R., Takeda K.

    *SEM 2022 - 11th Joint Conference on Lexical and Computational Semantics, Proceedings of the Conference     頁: 139 - 150   2022年

     詳細を見る

    記述言語:日本語   出版者・発行元:*SEM 2022 - 11th Joint Conference on Lexical and Computational Semantics, Proceedings of the Conference  

    There have been many successful applications of sentence embedding methods. However, it has not been well understood what properties are captured in the resulting sentence embeddings depending on the supervision signals. In this paper, we focus on two types of sentence embedding methods with similar architectures and tasks: one fine-tunes pre-trained language models on the natural language inference task, and the other fine-tunes pre-trained language models on word prediction task from its definition sentence, and investigate their properties. Specifically, we compare their performances on semantic textual similarity (STS) tasks using STS datasets partitioned from two perspectives: 1) sentence source and 2) superficial similarity of the sentence pairs, and compare their performances on the downstream and probing tasks. Furthermore, we attempt to combine the two methods and demonstrate that combining the two methods yields substantially better performance than the respective methods on unsupervised STS tasks and downstream tasks.

    Scopus

  13. Program Chairs’ Preface

    Bhattacharya A., Li J.L.M., Agrawal D., Deshpande P.M., Jiang D., Krishnamurthy R., Gupta R., Takeda K., Bellatreche L., Pudi V., Srinivasa S., Fournier-Viger P.

    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)   13247 LNCS 巻   2022年

     詳細を見る

    記述言語:日本語   出版者・発行元:Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)  

    Scopus

  14. Leveraging Three Types of Embeddings from Masked Language Models in Idiom Token Classification

    Takahashi R., Sasano R., Takeda K.

    *SEM 2022 - 11th Joint Conference on Lexical and Computational Semantics, Proceedings of the Conference     頁: 234 - 239   2022年

     詳細を見る

    記述言語:日本語   出版者・発行元:*SEM 2022 - 11th Joint Conference on Lexical and Computational Semantics, Proceedings of the Conference  

    Many linguistic expressions have idiomatic and literal interpretations, and the automatic distinction of these two interpretations has been studied for decades. Recent research has shown that contextualized word embeddings derived from masked language models (MLMs) can give promising results for idiom token classification. This indicates that contextualized word embedding alone contains information about whether the word is being used in a literal sense or not. However, we believe that more types of information can be derived from MLMs and that leveraging such information can improve idiom token classification. In this paper, we leverage three types of embeddings from MLMs; uncontextualized token embeddings and masked token embeddings in addition to the standard contextualized word embeddings and show that the newly added embeddings significantly improve idiom token classification for both English and Japanese datasets.

    Scopus

  15. Cross-Modal Similarity-Based Curriculum Learning for Image Captioning

    Zhang H., Sugawara S., Aizawa A., Zhou L., Sasano R., Takeda K.

    Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022     頁: 7599 - 7606   2022年

     詳細を見る

    出版者・発行元:Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, EMNLP 2022  

    Image captioning models require the high-level generalization ability to describe the contents of various images in words. Most existing approaches treat the image-caption pairs equally in their training without considering the differences in their learning difficulties. Several image captioning approaches introduce curriculum learning methods that present training data with increasing levels of difficulty. However, their difficulty measurements are either based on domain-specific features or prior model training. In this paper, we propose a simple yet efficient difficulty measurement for image captioning using cross-modal similarity calculated by a pretrained vision-language model. Experiments on the COCO and Flickr30k datasets show that our proposed approach achieves superior performance and competitive convergence speed to baselines without requiring heuristics or incurring additional training costs. Moreover, the higher model performance on difficult examples and unseen data also demonstrates the generalization ability.

    Scopus

  16. Automating Interlingual Homograph Recognition with Parallel Sentences

    Han Y., Sasano R., Takeda K.

    2nd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 12th International Joint Conference on Natural Language Processing - Findings of the Association for Computational Linguistics: AACL-IJCNLP 2022     頁: 211 - 216   2022年

     詳細を見る

    出版者・発行元:2nd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 12th International Joint Conference on Natural Language Processing - Findings of the Association for Computational Linguistics: AACL-IJCNLP 2022  

    Interlingual homographs are words that spellthe same but possess different meanings acrosslanguages. Recognizing interlingual homographs from form-identical words generallyneeds linguistic knowledge and massive annotation work. In this paper, we propose anautomatic interlingual homograph recognitionmethod based on the cross-lingual word embedding similarity and co-occurrence of formidentical words in parallel sentences. We conduct experiments with off-the-shelf languagemodels coordinating with cross-lingual alignment operations and co-occurrence metrics onthe Chinese-Japanese and English-Dutch language pairs. Experimental results demonstratethat our proposed method can achieve accurateand consistent predictions across languages.

    Scopus

  17. Zero-Shot Translation Quality Estimation with Explicit Cross-Lingual Patterns

    Zhou L., Ding L., Takeda K.

    5th Conference on Machine Translation, WMT 2020 - Proceedings     頁: 1068 - 1074   2021年

     詳細を見る

    記述言語:日本語   出版者・発行元:5th Conference on Machine Translation, WMT 2020 - Proceedings  

    This paper describes our submission of the WMT 2020 Shared Task on Sentence Level Direct Assessment, Quality Estimation (QE). In this study, we empirically reveal the mismatching issue when directly adopting BERTScore (Zhang et al., 2020) to QE. Specifically, there exist lots of mismatching errors between source sentence and translated candidate sentence with token pairwise similarity. In response to this issue, we propose to expose explicit cross lingual patterns, e.g. word alignments and generation score, to our proposed zero-shot models. Experiments show that our proposed QE model with explicit cross-lingual patterns could alleviate the mismatching issue, thereby improving the performance. Encouragingly, our zero-shot QE method could achieve comparable performance with supervised QE method, and even outperforms the supervised counterpart on 2 out of 6 directions. We expect our work could shed light on the zero-shot QE model improvement.

    Scopus

  18. Verb Sense Clustering using Contextualized Word Representations for Semantic Frame Induction

    Yamada K., Sasano R., Takeda K.

    Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021     頁: 4353 - 4362   2021年

     詳細を見る

    記述言語:日本語   出版者・発行元:Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021  

    Contextualized word representations have proven useful for various natural language processing tasks. However, it remains unclear to what extent these representations can cover hand-coded semantic information such as semantic frames, which specify the semantic role of the arguments associated with a predicate. In this paper, we focus on verbs that evoke different frames depending on the context, and we investigate how well contextualized word representations can recognize the difference of frames that the same verb evokes. We also explore which types of representation are suitable for semantic frame induction. In our experiments, we compare seven different contextualized word representations for two English frame-semantic resources, FrameNet and PropBank. We demonstrate that several contextualized word representations, especially BERT and its variants, are considerably informative for semantic frame induction. Furthermore, we examine the extent to which the contextualized representation of a verb can estimate the number of frames that the verb can evoke.

    Scopus

  19. Semantic Frame Induction using Masked Word Embeddings and Two-Step Clustering

    Yamada Kosuke, Sasano Ryohei, Takeda Koichi

    ACL-IJCNLP 2021: THE 59TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 11TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING, VOL 2     頁: 811 - 816   2021年

     詳細を見る

    記述言語:日本語  

    Web of Science

  20. Semantic Frame Induction using Masked Word Embeddings and Two-Step Clustering

    Yamada K., Sasano R., Takeda K.

    ACL-IJCNLP 2021 - 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Proceedings of the Conference   2 巻   頁: 811 - 816   2021年

     詳細を見る

    記述言語:日本語   出版者・発行元:ACL-IJCNLP 2021 - 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Proceedings of the Conference  

    Recent studies on semantic frame induction show that relatively high performance has been achieved by using clustering-based methods with contextualized word embeddings. However, there are two potential drawbacks to these methods: One is that they focus too much on the superficial information of the frameevoking verb and the other is that they tend to divide the instances of the same verb into too many different frame clusters. To overcome these drawbacks, we propose a semantic frame induction method using masked word embeddings and two-step clustering. Through experiments on the English FrameNet data, we demonstrate that using the masked word embeddings is effective for avoiding too much reliance on the surface information of frameevoking verbs and that two-step clustering can improve the number of resulting frame clusters for the instances of the same verb.

    Scopus

  21. Self-Guided Curriculum Learning for Neural Machine Translation

    Zhou, L; Ding, L; Duh, K; Watanabe, S; Sasano, R; Takeda, K

    IWSLT 2021: THE 18TH INTERNATIONAL CONFERENCE ON SPOKEN LANGUAGE TRANSLATION     頁: 206 - 214   2021年

     詳細を見る

    記述言語:日本語  

    Web of Science

  22. Driving Behavior Aware Caption Generation for Egocentric Driving Videos Using In-Vehicle Sensors

    Zhang, HK; Takeda, K; Sasano, R; Adachi, Y; Ohtani, K

    2021 IEEE INTELLIGENT VEHICLES SYMPOSIUM WORKSHOPS (IV WORKSHOPS)     頁: 287 - 292   2021年

     詳細を見る

    記述言語:日本語   出版者・発行元:IEEE Intelligent Vehicles Symposium, Proceedings  

    Video captioning aims to generate textual descriptions according to the video contents. The risk assessment of autonomous driving vehicles has become essential for an insurance company for providing adequate insurance coverage, in particular, for emerging MaaS business. The insurers need to assess the risk of autonomous driving business plans with a fixed route by analyzing a large number of driving data, including videos recorded by dash cameras and sensor signals. To make the process more efficient, generating captions for driving videos can provide insurers concise information to understand the video contents quickly. A natural problem with driving video captioning is, since the absence of egovehicles in these egocentric videos, descriptions of latent driving behaviors are difficult to be grounded in specific visual cues. To address this issue, we focus on generating driving video captions with accurate behavior descriptions, and propose to incorporate in-vehicle sensors which encapsulate the driving behavior information to assist the caption generation. We evaluate our method on the Japanese driving video captioning dataset called City Traffic, where the results demonstrate the effectiveness of in-vehicle sensors on improving the overall performance of generated captions, especially on generating more accurate descriptions for the driving behaviors.

    DOI: 10.1109/IVWorkshops54471.2021.9669259

    Web of Science

    Scopus

  23. DefSent: Sentence Embeddings using Definition Sentences

    Tsukagoshi, H; Sasano, R; Takeda, K

    ACL-IJCNLP 2021: THE 59TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS AND THE 11TH INTERNATIONAL JOINT CONFERENCE ON NATURAL LANGUAGE PROCESSING, VOL 2     頁: 411 - 418   2021年

     詳細を見る

    記述言語:日本語  

    Web of Science

  24. DefSent: Sentence Embeddings using Definition Sentences

    Tsukagoshi H., Sasano R., Takeda K.

    ACL-IJCNLP 2021 - 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Proceedings of the Conference   2 巻   頁: 411 - 418   2021年

     詳細を見る

    記述言語:日本語   出版者・発行元:ACL-IJCNLP 2021 - 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, Proceedings of the Conference  

    Sentence embedding methods using natural language inference (NLI) datasets have been successfully applied to various tasks. However, these methods are only available for limited languages due to relying heavily on the large NLI datasets. In this paper, we propose DefSent, a sentence embedding method that uses definition sentences from a word dictionary, which performs comparably on unsupervised semantics textual similarity (STS) tasks and slightly better on SentEval tasks than conventional methods. Since dictionaries are available for many languages, DefSent is more broadly applicable than methods using NLI datasets without constructing additional datasets. We demonstrate that DefSent performs comparably on unsupervised semantics textual similarity (STS) tasks and slightly better on SentEval tasks to the methods using large NLI datasets.

    Scopus

  25. Transformer-based Lexically Constrained Headline Generation

    Yamada K., Hitomi Y., Tamori H., Sasano R., Okazaki N., Inui K., Takeda K.

    EMNLP 2021 - 2021 Conference on Empirical Methods in Natural Language Processing, Proceedings     頁: 4085 - 4090   2021年

     詳細を見る

    記述言語:日本語   出版者・発行元:EMNLP 2021 - 2021 Conference on Empirical Methods in Natural Language Processing, Proceedings  

    This paper explores a variant of automatic headline generation methods, where a generated headline is required to include a given phrase such as a company or a product name. Previous methods using Transformer-based models generate a headline including a given phrase by providing the encoder with additional information corresponding to the given phrase. However, these methods cannot always include the phrase in the generated headline. Inspired by previous RNN-based methods generating token sequences in backward and forward directions from the given phrase, we propose a simple Transformer-based method that guarantees to include the given phrase in the high-quality generated headline. We also consider a new headline generation strategy that takes advantage of the controllable generation order of Transformer. Our experiments with the Japanese News Corpus demonstrate that our methods, which are guaranteed to include the phrase in the generated headline, achieve ROUGE scores comparable to previous Transformer-based methods. We also show that our generation strategy performs better than previous strategies.

    Scopus

  26. Self-Guided Curriculum Learning for Neural Machine Translation

    Zhou L., Ding L., Duh K., Watanabe S., Sasano R., Takeda K.

    IWSLT 2021 - 18th International Conference on Spoken Language Translation, Proceedings     頁: 206 - 214   2021年

     詳細を見る

    記述言語:日本語   出版者・発行元:IWSLT 2021 - 18th International Conference on Spoken Language Translation, Proceedings  

    In supervised learning, a well-trained model should be able to recover ground truth accurately, i.e. the predicted labels are expected to resemble the ground truth labels as much as possible. Inspired by this, we formulate a difficulty criterion based on the recovery degrees of training examples. Motivated by the intuition that after skimming through the training corpus, the neural machine translation (NMT) model “knows” how to schedule a suitable curriculum according to learning difficulty, we propose a self-guided curriculum learning strategy that encourages the NMT model to learn from easy to hard on the basis of recovery degrees. Specifically, we adopt sentence-level BLEU score as the proxy of recovery degree. Experimental results on translation benchmarks including WMT14 English?German and WMT17 Chinese?English demonstrate that our proposed method considerably improves the recovery degree, thus consistently improving the translation performance.

    Scopus

  27. 「いいね」「シェア」をした投稿のテキスト情報を利用したSNSユーザの性格推定 査読有り

    山田康輔, 笹野遼平, 武田浩一

    人工知能学会論文誌   35 巻 ( 4 ) 頁: 1 - 12   2020年7月

     詳細を見る

    記述言語:日本語   掲載種別:研究論文(学術雑誌)  

  28. テキスト分析のためのOLAPシステム

    猪口明博,武田浩一

    情報処理学会論文誌データベース   48 巻 ( SIG 11 ) 頁: 58 - 68   2007年6月

  29. MedTAKMI-CDI: Interactive knowledge discovery for clinical decision intelligence

    Akihiro Inokuchi, Koichi Takeda, Noriko Inaoka, Fumihiko Wakao

    IBM Systems Journal   46 巻 ( 1 ) 頁: 115 - 133   2007年1月

  30. A text-mining system for knowledge discovery from biomedical document

    Naohiko Uramoto, Hirofumi Matsuzawa, Tohru Nagano, Akiko Murakami, Hironori Tekeuchi , Koichi Takeda

    IBM Systems Journal   43 巻 ( 3 ) 頁: 516 - 533   2004年7月

  31. 知識ベースを利用した機械翻訳システムShalt2

    武田浩一, 浦本直彦, 那須川哲哉, 荻野紫穂, 堤泰次郎

    コンピュータソフトウェア   12 巻 ( 5 ) 頁: 22 - 32   1995年9月

  32. 日本語文書構成支援システムの設計と評価

    鈴木恵美子,武田浩一,

    情報処理学会論文誌   30 巻 ( 11 ) 頁: 1402 - 1412   1989年11月

  33. Analysis and generation grammars

    Donna Gates, Koichi Takeda, Teruko Mitamura, Lori S. Levin, Marion Kee

      4 巻 ( 1 ) 頁: 53 - 66   1989年3月

  34. Lexicons

    Donna Gates, Dawn Haberlach, Todd Kaufmann, Marion Kee, Rita McCardell, Teruko Mitamura, Ira Monarch, Stephen Morrisson, Sergei Nirenburg, Eric Nyberg, Koichi Takeda

      4 巻 ( 1 ) 頁: 67 - 112   1989年3月

  35. 統計的手法による漢字複合語の自動分割

    武田浩一, 藤崎哲之助

    情報処理学会論文誌   28 巻 ( 9 ) 頁: 952 - 961   1987年9月

▼全件表示

書籍等出版物 2

  1. インターネット機械翻訳の世界

    宮平知博 渡辺日出雄 田添英一 神山淑朗 武田浩一( 担当: 共著)

    毎日コミュニケーションズ  2000年2月  ( ISBN:4839903115

     詳細を見る

    総ページ数:190   記述言語:日本語

  2. The KBMT Project: A Case Study in Knowledge-based Machine Translation

    Sergei Nireburg, Ira Monarch, Todd Kaufmann, Lori Levin, Teruko Mitamura, Donna Gates, Koichi Takeda, Mario Kee, Margalit Zabludowski, Dawn Haberlach, Rita McCarddell, Stephen Morrisson, Ralf Brown, Eric Nyberg III( 担当: 共著)

    Morgan Kaufmann  1990年7月 

     詳細を見る

    記述言語:英語 著書種別:学術書

講演・口頭発表等 29

  1. 定義文を用いた文埋め込み構成法

    塚越駿, 笹野遼平, 武田浩一

    言語処理学会 第27回年次大会 

     詳細を見る

    開催年月日: 2021年3月

  2. マスクされた単語の埋め込みと2段階クラスタリングを用いた動詞の意味フレーム推定

    山田康輔, 笹野遼平, 武田浩一

    言語処理学会 第27回年次大会 

     詳細を見る

    開催年月日: 2021年3月

  3. 静的な単語埋め込みによるカタカナ語を対象としたBERTの語彙拡張

    平子潤, 笹野遼平, 武田浩一

    言語処理学会 第27回年次大会 

     詳細を見る

    開催年月日: 2021年3月

  4. Zero-Shot Translation Quality Estimation with Explicit Cross-Lingual Patterns 国際共著 国際会議

    Lei Zhou, Liang Ding, Koichi Takeda

    5th Conference on Machine Translation (WMT)  

     詳細を見る

    開催年月日: 2020年11月

    会議種別:口頭発表(一般)  

  5. Sequential Span Classification with Neural Semi-Markov CRFs for Biomedical Abstracts 国際会議

    Kosuke Yamada, Tsutomu Hirao, Ryohei Sasano, Koichi Takeda, Masaaki Nagata

    Findings of the Association for Computational Linguistics: EMNLP 2020 

     詳細を見る

    開催年月日: 2020年11月

  6. 文脈化単語埋め込みを用いた慣用句判定

    高橋 良輔, 笹野 遼平, 武田 浩一

    情報処理学会 第245回自然言語処理研究会 

     詳細を見る

    開催年月日: 2020年9月

  7. フレーム知識の自動獲得に向けた文脈化単語埋め込みの有用性の検証

    山田康輔, 笹野遼平, 武田浩一

    情報処理学会 第244回自然言語処理研究会 

     詳細を見る

    開催年月日: 2020年7月

  8. Development of a Medical Incident Report Corpus with Intention and Factuality Annotation 国際会議

    Hongkuan Zhang, Ryohei Sasano, Koichi Takeda, Zoie Shui-Yee Wong

    12th International Conference on Language Resources and Evaluation (LREC)  

     詳細を見る

    開催年月日: 2020年5月

    記述言語:英語   会議種別:口頭発表(一般)  

  9. Multilingualization of Question Answering Using Universal Dependencies 国際会議

    Hiroshi Kanayama, Koichi Takeda

    ACM SIGIR Workshop on OKBQA (Open Knowledge Base and Question-Answering) 

     詳細を見る

    開催年月日: 2017年8月

    記述言語:英語   会議種別:口頭発表(一般)  

    国名:日本国  

  10. Indexing Event Sequences for Medical Record Analysis 国際会議

    Daisuke Takuma, Yuta Tsuboi, Koichi Takeda, IsseiYoshida, Takahide Nogayama, Ai Ishida, Noriko Inaoka, and Yujiro Takai

    International Workshop on Pattern Recognition for Healthcare Analytics 

     詳細を見る

    開催年月日: 2012年11月

  11. An Online Analytical Processing of Text Data 国際会議

    Akihiro Inokuchi, Koichi Takeda

    16th Conf. on Information and Knowledge Management (CIKM) 

     詳細を見る

    開催年月日: 2007年11月

  12. Sentence generation for pattern-based machine translation 国際会議

    Koichi Takeda

    9th Conference on Theoretical and Methodological Issues (TMI) in Machine Translation 

     詳細を見る

    開催年月日: 2002年3月

  13. A Pattern-based Machine Translation System Extended by Example-based Processing 国際会議

    Hideo Watanabe, Koichi Takeda

    17th International Conference on Computational Linguistics 

     詳細を見る

    開催年月日: 1998年8月

  14. A Method for Relating Multiple Newspaper Articles by Using Graphs, and Its Application to Webcasting 国際会議

    Naohiko Uramoto, Koichi Takeda

    17th International Conference on Computational Linguistics 

     詳細を見る

    開催年月日: 1998年8月

  15. Site Outlining 国際会議

    Koichi Takeda

    3rd ACM Conf. on Digital Libraries 

     詳細を見る

    開催年月日: 1998年6月

  16. Pattern-Based Machine Translation 国際会議

    Koichi Takeda

    16th International Conference on Computational Linguistics (COLING) 

     詳細を見る

    開催年月日: 1996年8月

  17. Pattern-Based Context-Free Grammars for Machine Translation 国際会議

    Koichi Takeda

    34th Annual Meeting of the Association of Computational Linguistics (ACL) 

     詳細を見る

    開催年月日: 1996年7月

  18. Portable Knowledge Sources for Machine Translation 国際会議

    Koichi Takeda

    15th International Conference on Computational Linguistics (COLING) 

     詳細を見る

    開催年月日: 1994年8月

  19. Tricolor DAG's for Machine Translation" 国際会議

    Koichi Takeda

    32nd Annual Meeting of the Association of Computational Linguistics (ACL) 

     詳細を見る

    開催年月日: 1994年6月

  20. 精神疾患の診断補助のための自伝的記憶の詳細度による分類

    大柳慶悟, 武田浩一, 笹野遼平, ハルフォード デイビッド, 高野慶輔

    情報処理学会 第250回自然言語処理研究会  2021年9月28日 

  21. Automatic Interlingual Homograph Recognition with Context Features

    Yi Han, Ryohei Sasano, Koichi Takeda

    2022年3月17日 

  22. 直近1年の動向を考慮した最新論文のインパクト予測

    平子潤, 笹野遼平, 武田浩一

    言語処理学会 第28回年次大会  2022年3月16日 

  23. 自然言語推論と再現器を用いたSplit and Rephrase における生成文の品質向上

    塚越駿, 平尾努, 森下睦, 帖佐克己, 笹野遼平, 武田浩一

    言語処理学会 第28回年次大会  2022年3月16日 

  24. Transformer-based Lexically Constrained Headline Generation 国際会議

    Kosuke Yamada, Yuta Hitomi, Hideaki Tamori, Ryohei Sasano, Naoaki Okazaki, Kentaro Inui, Koichi Takeda

    2021 Conference on Empirical Methods in Natural Language Processing (EMNLP 2021)  2021年11月8日 

  25. Self-Guided Curriculum Learning for Neural Machine Translation 国際会議

    Lei Zhou, Liang Ding, Kevin Duh, Shinji Watanabe, Ryohei Sasano, Koichi Takeda

    18th International Conference on Spoken Language Translation (IWSLT 2021)  2021年8月6日 

  26. Semantic Frame Induction using Masked Word Embeddings and Two-Step Clustering 国際会議

    Kosuke Yamada, Ryohei Sasano, Koichi Takeda

    Joint Conference of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (ACL-IJCNLP 2021)  2021年8月4日 

  27. DefSent: Sentence Embeddings using Definition Sentences 国際会議

    Hayato Tsukagoshi, Ryohei Sasano, Koichi Takeda

    Joint Conference of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (ACL-IJCNLP 2021)  2021年8月3日 

  28. Verb Sense Clustering Using Contextualized Word Representations for Semantic Frame Induction

    Kosuke Yamada, Ryohei Sasano, Koichi Takeda

    Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021  2021年8月2日 

  29. Driving Behavior Aware Caption Generation for Egocentric Driving Videos Using In-Vehicle Sensors

    Hongkuan Zhang, Koichi Takeda, Ryohei Sasano, Yusuke Adachi, Kento Ohtani

    2021 IEEE Intelligent Vehicles Symposium Workshops (IV Workshops)  2021年7月11日 

▼全件表示

共同研究・競争的資金等の研究課題 1

  1. 人工知能技術適用によるスマート社会の実現/ 人工知能技術の社会実装に関する日米共同研究開発/ 判断根拠を言語化する人工知能の研究開発 国際共著

    2018年5月 - 2023年2月

    次世代人工知能・ロボット中核技術開発 

      詳細を見る

    担当区分:研究分担者  資金種別:競争的資金

産業財産権 1

  1. View composition system for synthesizing multiple visual representations of multiple data sets

    Koichi Takeda

     詳細を見る

    出願人:International Business Machines Corporation

    特許番号/登録番号:6,867,788  登録日:2005年3月 

    出願国:外国  

 

メディア報道 6

  1. アドヴィックス:名古屋大学と産学提携し、新世代のAI実用化のため、次世代ブレーキ開発を加速 インターネットメディア

    株式会社 三栄  モーターファンテック  2021年6月

  2. 「ワトソン」は救世主か? 新聞・雑誌

    週刊東洋経済  2015年5月

  3. IBM Watsonプロジェクトは「医療」から始まった, インターネットメディア

    日経デジタルヘルス  2015年3月

  4. コンピューターが進歩すると仕事がなくなる? インターネットメディア

    日経プラスワン  2014年3月

  5. 2020年クラウド時代のサバイバル特集 新聞・雑誌

    日経コミュニケーション  2014年1月

  6. ロボットVS職人社長 新聞・雑誌

    日経ビジネス  2013年8月

▼全件表示