研究者詳細 - 松原　茂樹

論文 - 松原　茂樹

分割表示 235 件中 61 - 80 件目／全件表示 >>

Parsing gapping constructions based on grammatical and semantic roles

Kato Y., Matsubara S.

EMNLP 2020 - 2020 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference 頁： 2747 - 2752 2020年

記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：EMNLP 2020 - 2020 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference

A gapping construction consists of a coordinated structure where redundant elements are elided from all but one conjuncts. This paper proposes a method of parsing sentences with gapping to recover elided elements. The proposed method is based on constituent trees annotated with grammatical and semantic roles that are useful for identifying elided elements. Our method outperforms the previous method in terms of F-measure and recall.

Scopus

PTB Graph Parsing with Tree Approximation 査読有り

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL-2019) 頁： 5344-5349 2019年7月

同時通訳における作業記憶の負荷と訳出との関係査読有り

笠浩一朗, 松原茂樹

通訳翻訳研究 ( 18 ) 頁： 1-12 2019年2月

同時通訳における語の欠落に影響を及ぼす要因の分析査読有り

蔡仲煕, 笠浩一朗, 松原茂樹

通訳翻訳研究 ( 18 ) 頁： 1-14 2019年2月

Statistical analyses of missing translations in simultaneous interpretation using a large-scale bilingual speech corpus

Cai Z., Ryu K., Matsubara S.

LREC 2018 - 11th International Conference on Language Resources and Evaluation 頁： 4282 - 4288 2019年

記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：LREC 2018 - 11th International Conference on Language Resources and Evaluation

This paper describes statistical analyses of missing translations in simultaneous interpretations. Eighty-eight lectures from English-to-Japanese interpretation data from a large-scale bilingual speech corpus were used for the analyses. Word-level alignment was provided manually, and English words without corresponding Japanese words were considered missing translations. The English lectures contained 46,568 content words, 33.1% of which were missing in the translation. We analyzed the relationship between missing translations and various factors, including the speech rate of the source language, delay of interpretation, part-of-speech, and depth in the syntactic structure of the source language. The analyses revealed that the proportion of missing translations is high when the speech rate is high and delay is large. We also found that a high proportion of adverbs were missed in the translations, and that words at deeper positions in the syntactic structure were more likely to be missed.

Scopus

PTB Graph Parsing with Tree Approximation

Kato, Y; Matsubara, S

57TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2019) 頁： 5344 - 5349 2019年

Model-Theoretic Incremental Interpretation Based on Discourse Representation Theory 査読有り

Yoshihide Kato, Shigeki Matsubara

Proceedings of the 32nd Pacific Asia Conference on Language, Information and Computation (PACLIC 32) 頁： 1 2018年12月

Statistical Analysis of Missing Translation in Simultaneous Interpretation Using A Large-scale Bilingual Speech Corpus 査読有り

Zhongxi Cai, Koichiro Ryu, Shigeki Matsubara

Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018) 頁： 4282-4288 2018年5月

語りの傾聴を話し手に示す応答発話の収集

村田匡輝, 大野誠寛, 松原茂樹

電気学会論文誌Ｃ（電子・情報・システム部門誌） 138 巻 ( 5 ) 頁： 637 - 638 2018年5月

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：一般社団法人電気学会

This paper describes collection of responsive utterances of showing attentive hearing for narrative speech. In this research, a method of annotating an existing narrative corpus with responsive utterances was adopted. Actually, 16,213 responsive utterances were collected for a narrative corpus consisting of 13,381 utterances. It was confirmed that the method can collect a variety of natural responsive utterances.

DOI： 10.1541/ieejeiss.138.637

Scopus

CiNii Research

同時通訳における語の欠落に影響を及ぼす要因の分析

蔡仲煕, 笠浩一朗, 松原茂樹

通訳翻訳研究 18 巻 ( 0 ) 頁： 133 - 146 2018年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：日本通訳翻訳学会

This paper describes corpus-based analyses of missing translations in simultaneous interpretations. We used English-to-Japanese interpretation data in the simultaneous interpretation database of Nagoya University for the analyses. The English lectures contained 46,568 content words, and 33.1% of them were missing in the translation. We analyzed the relations between omissions and the factors that influence them, i.e., the speech rate of the source language, delay of interpretation, part-of-speech, and depth in the syntactic structure of the source language. The analyses revealed that the proportion of missing translations is high when the speech rate is high, and the delay of interpretation is large, i.e., adverbs were missing in the translations in a high proportion; the words at deeper positions in the syntactic structure were more likely to be missed.

DOI： 10.50837/its.1808

CiNii Research

同時通訳における作業記憶の負荷と訳出との関係国際誌

笠浩一朗, 松原茂樹

通訳翻訳研究 18 巻 ( 0 ) 頁： 147 - 158 2018年

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：日本通訳翻訳学会

By employing corpus-based analyses of English-Japanese simultaneous interpretation, we discuss the influence translations have on the working memory load of simultaneous interpreters. To conduct a quantitative analysis with a large corpus, we used the Simultaneous Interpretation Database of Nagoya University. We analyzed a total of 88 sets of simultaneous interpretations of 22 different lectures in English given by 17 interpreters. When the load on the working memory of simultaneous interpreters is increased, the speech time increases, the translation rate decreases, but the average speech rate does not exhibit a relevant change.

DOI： 10.50837/its.1809

CiNii Research

Model-theoretic incremental interpretation based on discourse representation theory

Kato Y., Matsubara S.

Proceedings of the 32nd Pacific Asia Conference on Language, Information and Computation, PACLIC 2018 頁： 254 - 262 2018年

記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：Proceedings of the 32nd Pacific Asia Conference on Language, Information and Computation, PACLIC 2018

This paper proposes a model-theoretic approach to incremental interpretation where allsentence prefixes have semantic values. Theproposed semantics is based on DiscourseRepresentation Theory (DRT), where semantic representations (called DRSs) are interpreted as assignment updates. In our semantics, a partial DRS of a sentence prefix is interpreted as two sets which stipulate the assignment updates. One denotes possible updatesand the other denotes necessary updates. Withthe proposed semantics, we can assign truthvalues to sentence prefixes.

Scopus

Statistical Analyses of Missing Translations in Simultaneous Interpretation Using a Large-scale Bilingual Speech Corpus

Cai, ZX; Ryu, K; Matsubara, S

PROCEEDINGS OF THE ELEVENTH INTERNATIONAL CONFERENCE ON LANGUAGE RESOURCES AND EVALUATION (LREC 2018) 頁： 4282 - 4288 2018年

Correcting Syntactic Annotation Errors Based on Tree Mining 査読有り

Kanta Suzuki, Yoshihide Kato, Shigeki Matsubara

IEICE Transactions on Information and Systems E100-D 巻 ( 5 ) 頁： 1106-1113 2017年5月

Collection of responsive utterances to show attentive hearing attitude to speakers

Ohno T., Murata M., Matsubara S.

Proceedings of the 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017 ( 31 ) 2017年1月

記述言語：日本語掲載種別：研究論文（学術雑誌）出版者・発行元：Proceedings of the 11th International Conference on Ubiquitous Information Management and Communication, IMCOM 2017

One of the stressors for users who have a dialogue with a robot is a doubt about whether or not the robot can comprehend user's utterances. To solve this problem, it is effective for a robot to actively show the attitude of attentive hearing for user's utterances. This paper describes collection of responsive utterances to show the attitude of attentive hearing for the purpose of actualizing conversation robots with attitudes of attentive hearing. In the data collection, we made a worker respond to utterances in Japanese elders' narrative corpus so that the responses become appropriate as a role of a listener. As the result, we collected 4,885 responsive utterances uttered for narratives of 9 elders.

DOI： 10.1145/3022227.3022257

Scopus

対話コーパスを用いた相づち生成タイミングの検出査読有り

大野誠寛，神谷優貴，松原茂樹

電子情報通信学会論文誌 J100-A 巻 ( 1 ) 頁： 53-65 2017年1月

Left-Corner Parsing for Identifying PTB-Style Nonlocal Dependencies

Kato Yoshihide, Matsubara Shigeki

自然言語処理 24 巻 ( 3 ) 頁： 371 - 394 2017年

記述言語：英語掲載種別：研究論文（学術雑誌）出版者・発行元：一般社団法人　言語処理学会

<p>Nonlocal dependencies represent syntactic phenomenon such as wh-movement, A-movement in passives, topicalization, raising, control, and right node raising. Nonlocal dependencies play an important role in semantic interpretation. This paper proposes a left-corner parser that identifies nonlocal dependencies. Our parser integrates nonlocal dependency identification into a transition-based system. We adopt a left-corner strategy in order to use the syntactic relation c-command, which plays an important role in nonlocal dependency identification. To utilize the global features captured by nonlocal dependencies, our parser uses a structured perceptron. In experimental evaluations, our parser achieved a good balance between constituent parsing and nonlocal dependency identification. </p>

DOI： 10.5715/jnlp.24.371

CiNii Research

Incremental Semantic Construction Based on Combinatory Categorial Grammar 査読有り

Yoshihide Kato, Shigeki Matsubara

IEICE Transactions on Information and Systems E99-D 巻 ( 9 ) 頁： 2368-2376 2016年9月

Transition-Based Left-Corner Parsing for Identifying PTB-Style Nonlocal Dependencies 査読有り

Yoshihide Kato, Shigeki Matsubara

Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (ACL-2016) 頁： 930-940 2016年8月

係り受け解析との同時実行に基づく日本語文の語順整序査読有り

大野誠寛，吉田和史，加藤芳秀，松原茂樹

電子情報通信学会論文誌 J99-D 巻 ( 2 ) 頁： 201-213 2016年2月

記述言語：日本語掲載種別：研究論文（学術雑誌）

日本語は語順が比較的自由であるといわれているが，語順に関する選好がないわけではない．そのため，文法的には間違っていないものの読みにくい語順をもった文が作成されることがある．本論文では，読みにくい語順の日本語文に対して，読みやすい語順となるように文節を並べ替える手法を提案する．本手法は，係り受け構造が付与されていない文を入力とし，係り受け解析と語順整序を同時に行う．係り受けと語順の適切さを同時に考慮することにより，読みやすい語順に精度よく整えることができる．評価実験では，文法的には間違っていないものの読みにくい語順をもつ文を新聞記事から擬似的に552文作成し，それらに対して語順整序を実行した．元の新聞記事文の語順を正解としたときの，語順整序結果との一致を測定した結果，本手法による語順整序結果は，比較のために設定した二つのベースラインと比べ高い正解率（83.82%）を達成しており，本手法の有効性を確認した．

　前のページ - 次のページ　

<松原 茂樹>

論文 - 松原 茂樹

<松原　茂樹>

論文 - 松原　茂樹