基於預訓練模型的抽取式文件摘要之研究｜國立臺灣科技大學博碩士論文系統

簡易檢索 / 詳目顯示

回結果列表

研究生：	吳政育 Zheng-Yu Wu
論文名稱：	基於預訓練模型的抽取式文件摘要之研究 Extractive Document Summarization Using Pre-trained Language Models
指導教授：	陳冠宇 Kuan-Yu Chen
口試委員:	陳柏琳 Ber-Lin Chen 鮑興國 Hsing-Kuo Pao
學位類別：	碩士 Master
系所名稱：	電資學院 - 資訊工程系 Department of Computer Science and Information Engineering
論文出版年：	2020
畢業學年度：	108
語文別：	中文
論文頁數：	105
中文關鍵詞：	摘要、抽取式、BERT 、強化學習、最大邊緣相關性
外文關鍵詞：	Summarization, Extractive, BERT, Reinforcement Learning, MMR
相關次數：	點閱：365 下載：0
分享至:	分享至facebook 分享至twitter

查詢本校圖書館目錄查詢臺灣博碩士論文知識加值系統勘誤回報

上一筆

目前主要的自動摘要方法，可以分為抽取式摘要(extractive)與重寫式摘要
(abstractive)。在本文中，我們提出無監督學習以及監督學習且基於 BERT 的摘要
任務架構，首先，我們提出一套資訊擴增之抽取式摘要模型 (Generating Pseudo
relevant Representations for Summarization, PRMSUM)，使用生成對抗網路生成偽
向量，用以補足文章中句子的資訊。接著我們提出一套新穎的次模函數之抽取式
摘要模型 (Submodular-based Extractive Summarization, SUBMODULARSUM)，不
僅考慮了次模函數在抽取式摘要上的應用，更直接地讓神經網路學習次模函數的
特性並且讓神經網路代替貪婪演算法選擇摘要句子。接著我們提出覆蓋率之抽取
式摘要模型 (Coverage-based Extractive Summarization, COVERSUM)，直接將覆
蓋率考慮至模型中，讓模型學習如何選擇有效的句子以覆蓋整個文章。最後，我
們提出基於 BERT 的強健性抽取式摘要方法 (Enhanced BERT-based Extractive
Summarization Framework, EBSUM)，它不僅考慮了句子的位置資訊、利用強化學
習增強摘要模型與評估標準的關聯性，更直接地將最大邊緣相關性(maximal
marginal relevance, MMR)概念融入摘要模型之中，以避免冗餘資訊的選取。

Automatic summarization methods can be classified into two major spectrums：
extractive summarization and abstractive summarization. In this paper, we propose supervised learning and unsupervised learning summarization architecture, and both of them are BERT-based. First, we propose a model of extractive summarization tasks with pseudo-vector representation using a generative adversarial network (PRMSUM). Not only the pseudo-vectors are considered and increase the amount of information of sentences in the document, we propose a new model of BERT-based unsupervised extractive summarization architecture (SUBMODULARSUM), which not only considers the effect of submodular functions on extractive summarization, but also lets neural networks learn the characteristics of the submodule function. Finally using the neural network selects sentences instead of using greedy algorithm. Then we propose the unsupervised learning model COVERSUM based on the coverage, which directly considers the coverage into the model and allows the model to learn how to choose sentences that cover the entire document. Finally, we propose enhanced BERT-based extractive summarization framework (EBSUM), which not only takes sentence position information and RL training into account, but the maximal marginal relevance (MMR) criterion is also considered.

內容
摘要  .................................................................................................................... 4
Abstract ................................................................................................................... 5
第 1 章 緒論 ............................................................................................................ 6
1 自然語言處理的簡介 ................................................................................ 6
2 自然語言處理的應用 ................................................................................ 7
3 自動摘要的研究現進展 ............................................................................ 7
4 研究目的及動機 ...................................................................................... 10
第 2 章 相關研究 .................................................................................................. 13
1 預訓練詞向量 .......................................................................................... 13
1.1 分散式向量表示法 .................................................................................. 13
1.2 Transformer .............................................................................................. 16
1.3 ELMo (Embedding from Language Models) .......................................... 18
1.4 OPENAI GPT .......................................................................................... 20
1.5 BERT(Bidirectional Encoder Representation from Transformers) ......... 21
2 重寫式摘要 .............................................................................................. 23
2.1 序列至序列模型 (Sequence-to-sequence Model) .................................. 23
2.2 指針生成網路 (Pointer-Generator Networks) ....................................... 25
2.3 加強式快速重寫式摘要模型 .................................................................. 28
2.4 生成對抗網路之重寫式摘要模型 .......................................................... 31
3 抽取式摘要 .............................................................................................. 32
3.1 基於遞歸神經網路之抽取式摘要模型 (Recurrent Neural Network based Sequence Model, Summarunner) ............................................................... 32
3.2 基於強化學習之抽取式摘要模型 (REinFoRcement Learning-based Summarization Model, Refresh) ........................................................................... 34
4 基於 BERT 的摘要方法 (BERT-based Summarization Methods)........ 35
4.1 BERTSUM ............................................................................................... 35
4.2 基於單句和雙句之重寫式摘要模型 ...................................................... 37
2

4.3 基於位置的強健式摘要模型(Position-Augmented Centrality based Summarization, PACSUM) .................................................................................. 39
5 資料檢索 .................................................................................................. 41
5.1 Rocchio’s 演算法 (Rocchio’s Algorithm) .............................................. 46
5.2 偽相關表示法模型 (Pseudo-relevant Representation Model, PRM) .... 48
第 3 章 偽相關表示法模型於摘要的應用 .......................................................... 51
1 動機以及目的 .......................................................................................... 51
2 PRMSUM 架構 ....................................................................................... 51
3 小結 .......................................................................................................... 53
第 4 章 次模函數之抽取式摘要模型 (Submodular-based Extractive Summarization Method, SUBMODULARSUM) ................................................ 54
1 動機以及目的 .......................................................................................... 54
2 次模函數 .................................................................................................. 54
3 句子以及文章表示法 .............................................................................. 56
4 次模函數評分模型 .................................................................................. 56
5 自動摘要模型 .......................................................................................... 57
6 小結 .......................................................................................................... 59
第 5 章 覆蓋率之抽取式摘要模型 (Coverage-based Extractive Summarization Method, COVERSUM) ........................................................................................ 60
1 動機以及目的 .......................................................................................... 60
2 最大邊緣相關性資訊 .............................................................................. 60
3 預訓練 ...................................................................................................... 62
4 COVERSUM 架構 .................................................................................. 63
5 強化學習 .................................................................................................. 64
6 小結 .......................................................................................................... 64
第 6 章 基於 BERT 的強健性抽取式摘要方法 (Enhanced BERT-based Extractive Summarization Framework, EBSUM) ................................................ 65
1 動機以及目的 .......................................................................................... 65
3

2 EBSUM 架構 ........................................................................................... 66
3 EBSUM 模型中的 BERT 句子表示法 ................................................... 67
4 句子位置向量 .......................................................................................... 67
5 強化學習 .................................................................................................. 69
6 最大邊緣相關性向量 .............................................................................. 70
7 微調抽取式摘要任務之模型 .................................................................. 71
8 小結 .......................................................................................................... 72
第 7 章 實驗 .......................................................................................................... 73
1 實驗語料 .................................................................................................. 73
2 評估方式 .................................................................................................. 74
3 實驗細節和基礎系統 .............................................................................. 75
4 實驗結果 .................................................................................................. 77
4.1 偽相關表示法模型於摘要的應用 .......................................................... 77
4.2 次模函數之抽取式摘要模型 .................................................................. 79
4.3 覆蓋率之抽取式摘要模型 ...................................................................... 81
4.4 基於 BERT 的強健性抽取式摘要方法 ................................................. 84
5 綜合結果 .................................................................................................. 90
第 8 章 結論 .......................................................................................................... 95
參考文獻 .............................................................................................................. 96 
                                

[1] "Deep contextualized word representations," https://zhuanlan.zhihu.com/p/54700412.
[2] Y.-C. Chen and M. Bansal, "Fast abstractive summarization with reinforceselected sentence rewriting," arXiv preprint arXiv:1805.11080, 2018.
[3] Y.-S. Wang and H.-Y. Lee, "Learning to encode text as human-readable summaries using generative adversarial networks," arXiv preprint arXiv:1810.02851, 2018.
[4] R. Nallapati et al., "Summarunner: a recurrent neural network based sequence model for extractive summarization of documents," in Association for the Advancement of Artificial Intelligence Conference on Artificial Intelligence, 2017.
[5] A. Radford et al.,"Improving language understanding by generative pretraining," https://s3-us-west2.amazonaws.com/openaiassets/researchcovers/languageunsupervised/languag e understandingpaper.pdf, 2018.
[6] K. C. Yang, " 從零開始的 Sequence to sequence," http://zake7749.github.io/2017/09/28/Sequence-to-Sequence-tutorial/.
[7] Z.-Y. Wu et al., "Generating pseudo-relevant representations for spoken document retrieval," in IEEE Acoustics, Speech and Signal Processing 2019.
[8] A. See et al., "Get to the point: summarization with pointer-generator networks," in Annual Meeting of the Association for Computational Linguistics, 2017.
[9] L. Lebanoff et al., "Scoring sentence singletons and pairs for abstractive summarization," in Annual Meeting of the Association for Computational Linguistics, 2019.
[10] S. Narayan et al., "Ranking sentences for extractive summarization with reinforcement learning," in North American Chapter of the Association for Computational Linguistics, 2018.
[11] E. Brill, "A simple rule-based part of speech tagger," in Applied natural language processing, 1992.
[12] Y. Zhang et al., "An extensible framework for internet booking application based on rule engine," in Web Information Systems and Applications Conference, 2009.
[13] K. J. Leonard, "The development of a rule based expert system model for fraud alert in consumer credit," in Journal of Operational Research,1995.
[14] J. Clark et al., "A neural network based approach to automated e-mail classification," in IEEE Web Intelligence., 2003.
[15] A. Secker et al., "AISEC: an artificial immune system for e-mail classification," in Congress on Evolutionary Computation, 2003. CEC'03., 2003. [16] W. Awad and S. ELseuofi, "Machine learning methods for spam e-mail classification," in Journal of Computer Science Information Technology, 2011. [17] N. O. F. Elssied et al., "Research article a novel feature selection based on oneway anova f-test for e-mail spam classification," in Journal of Applied Sciences, Engineering Technology, 2014.
[18] P. Pantel and D. Lin, "Spamcop: s spam classification & organization program," in Association for the Advancement of Artificial Intelligence-98 Workshop on Learning for Text Categorization, 1998.
[19] J. Alspector et al., "Group based spam classification," ed: Google Patents, 2010.
[20] I. Idris, "E-mail spam classification with artificial neural network and negative selection algorithm," in Journal of Computer Science Communication Networks, 2011.
[21] J. Huang et al., "Extracting chatbot knowledge from online discussion forums," in International Joint Conference on Artificial Intelligence, 2007. [22] S. A. Abdul-Kader and J. Woods, "Survey on chatbot design techniques in speech conversation systems," in Journal of Advanced Computer Science Application, 2015.
[23] B. A. Shawar and E. Atwell, "Different measurements metrics to evaluate a chatbot system," in workshop on bridging the gap: Academic and industrial research in dialog technologies, 2007.
[24] I. V. Serban et al., "A deep reinforcement learning chatbot," arXiv preprint arXiv:1709.02349, 2017.
[25] X. Hu and H. Liu, "Text analytics in social media," in Mining text data: Springer, 2012.
[26] D. Zeng et al., "Social media analytics and intelligence," in Journal of Intelligent Systems, 2010.
[27] S. Stieglitz et al., "Social media and political communication: a social media analytics framework," in Journal of Social network analysis and mining, 2013.
[28] B. Batrinca and P. C. Treleaven, "Social media analytics: a survey of techniques, tools and platforms," in Journal of Ai Society, 2015.
[29] N. J. Conroy et al., "Automatic deception detection: Methods for finding fake news," in Journal of Association for Information Science Technology, 2015. [30] K. Shu et al., "Fake news detection on social media: A data mining perspective," in Journal of Explorations Newsletter, 2017.
[31] N. Ruchansky et al., "Csi: A hybrid deep model for fake news detection," in Information and Knowledge Management, 2017.
[32] E. Tacchini et al., S. Moret, and L. de Alfaro, "Some like it hoax: Automated fake news detection in social networks," arXiv preprint arXiv:.07506, 2017.
[33] V. Pérez-Rosas et al., "Automatic detection of fake news," arXiv preprint arXiv:.07104, 2017.
[34] R. Mihalcea, "Language independent extractive summarization," in Artificial Intelligence, 2005.
[35] S. Narayan et al., "Ranking sentences for extractive summarization with reinforcement learning," arXiv preprint arXiv:1802.08636, 2018.
[36] Y. Liu, "Fine-tune bert for extractive summarization," arXiv preprint arXiv:1903.10318, 2019.
[37] Y. Gu and Y. Hu, "Extractive summarization with very deep pretrained language model," in Journal of Artificial Intelligence and Applications, 2019. [38] K.-F. Wong et al., "Extractive summarization using supervised and semisupervised learning," in Annual Meeting of the Association for Computational Linguistics, 2008.
[39] D. Parveen et al., "Topical coherence for graph-based extractive summarization," in Empirical Methods in Natural Language Processing, 2015.
[40] L. Liu et al., "Generative adversarial network for abstractive text summarization," in Association for the Advancement of Artificial Intelligence conference on artificial intelligence, 2018.
[41] R. Nallapati et al., "Abstractive text summarization using sequence-to-sequence rnns and beyond," arXiv preprint arXiv:1602.06023, 2016.
[42] F. Liu et al., "Toward abstractive summarization using semantic representations," arXiv preprint arXiv:1805.10399, 2018.
[43] S. Gehrmann et al., "Bottom-up abstractive summarization," arXiv preprint arXiv:1808.10792, 2018.
[44] L. Lebanoff et al., "Scoring sentence singletons and pairs for abstractive summarization," arXiv preprint arXiv:1906.00077, 2019.
[45] R. Paulus et al., "A deep reinforced model for abstractive summarization," arXiv preprint arXiv:1705.04304, 2017.
[46] R. Mihalcea, "Graph-based ranking algorithms for sentence extraction, applied to text summarization," in Association for Computational Linguistics, 2004.
[47] G. Erkan and D. R. Radev, "Lexrank: graph-based lexical centrality as salience in text summarization," in Journal of artificial intelligence research, 2004.
[48] M. Qiu et al., "Alime chat: a sequence to sequence and rerank based chatbot engine," in Annual Meeting of the Association for Computational Linguistics, 2017.
[49] R. J. Weiss et al., "Sequence-to-sequence models can directly translate foreign speech," arXiv preprint arXiv:.08581, 2017.
[50] C.-W. Lee et al., "Scalable sentiment for sequence-to-sequence chatbot response with performance analysis," in IEEE Acoustics, Speech and Signal Processing 2018.
[51] K. Dave et al., "Mining the peanut gallery: Opinion extraction and semantic classification of product reviews," in World Wide Web, 2003.
[52] A. Caramazza and I. Brones, "Semantic classification by bilinguals," in Journal of Psychology, 1980.
[53] X. Liu et al., "Representation learning using multi-task deep neural networks for semantic classification and information retrieval," in North American Chapter of the Association for Computational Linguistics, 2015.
[54] C.-Y. Lin, "Rouge: A package for automatic evaluation of summaries," in Annual Meeting of the Association for Computational Linguistics-04 Workshop, 2004.
[55] J. Devlin et al., "Bert: pre-training of deep bidirectional transformers for language understanding," in North American Chapter of the Association for Computational Linguistics, 2018.
[56] P. Shi and J. Lin, "Simple bert models for relation extraction and semantic role labeling," arXiv preprint arXiv:.05255, 2019.
[57] R. Nogueira and K. Cho, "Passage re-ranking with BERT," arXiv preprint arXiv:1901.04085, 2019.
[58] F. Souza et al., "Portuguese named entity recognition using BERT-CRF," arXiv preprint arXiv:.10649, 2019.
[59] T. Moon et al., "Towards lingua franca named entity recognition with BERT," arXiv preprint arXiv:.01389, 2019.
[60] W. Yang et al., "Simple applications of bert for ad hoc document retrieval," arXiv preprint arXiv:1903.10972, 2019.
[61] Z.-Y. Wu et al., "Generating pseudo-relevant representations for spoken document retrieval," in IEEE Acoustics, Speech and Signal Processing, 2019.
[62] A. Rambaut et al., "Posterior summarization in bayesian phylogenetics using tracer 1.7," in Journal of Systematic biology, 2018.
[63] Y. Billawala et al., "Scalable and effective document summarization framework," ed: Google Patents, 2018.
[64] L. A. Leiva, "Responsive snippets: adaptive skim-reading for mobile devices," in Human-Computer Interaction with Mobile Devices and Services Adjunct, 2018.
[65] M. T. Maybury and A. E. Merlino Jr, "Automated segmentation, information extraction, summarization, and presentation of broadcast news," ed: Google Patents, 2005.
[66] S. R. Maskey and J. Hirschberg, "Automatic summarization of broadcast news using structural features," in Eighth European Speech Communication and Technology, 2003.
[67] S. Stieglitz and L. Dang-Xuan, "Social media and political communication: a social media analytics framework," in Journal of Social network analysis and mining, 2013.
[68] W. Fan and M. D. Gordon, "The power of social media analytics," in Journal of Commun. Acm, 2014.
[69] M. Imran et al., "Processing social media messages in mass emergency: A survey," in Journal of ACM Computing Surveys, 2015.
[70] W. He et al.,"A novel social media competitive analytics framework with sentiment benchmarks," in Journal of Information & Management, 2015.
[71] Y. Goldberg and O. Levy, "Word2vec explained: deriving Mikolov et al.'s negative-sampling word-embedding method," arXiv preprint arXiv:1402.3722, 2014.
[72] T. Mikolov et al., "Efficient estimation of word representations in vector space," arXiv preprint arXiv:1301.3781, 2013.
[73] J. Pennington et al., "Glove: global vectors for word representation," in Empirical Methods in Natural Language Processing & International Joint Conference on Natural Language Processing, 2014.
[74] T. Mikolov et al., "Distributed representations of words and phrases and their compositionality," in Advances in neural information processing systems, 2013.
[75] D. Kiela et al., "Specializing word embeddings for similarity or relatedness," in Empirical Methods in Natural Language Processing & International Joint Conference on Natural Language Processing, 2015.
[76] M. Faruqui et al., "Retrofitting word vectors to semantic lexicons," arXiv preprint arXiv:1411.4166, 2014.
[77] M. Ono et al.,"Word embedding-based antonym detection using thesauri and distributional information," in North American Chapter of the Association for Computational Linguistics, 2015.
[78] R. Fu et al., "Learning semantic hierarchies via word embeddings," in Annual Conference of the Association for Computational Linguistics, 2014.
[79] L. Tan et al., "Usaar-wlv: Hypernym generation with deep neural nets," in International Workshop on Semantic Evaluation, 2015.
[80] A. Vaswani et al., "Attention is all you need," in Advances in neural information processing systems, 2017.
[81] M. E. Peters et al., "Deep contextualized word representations," arXiv preprint arXiv:1802.05365, 2018.
[82] R. Jozefowicz et al., "Exploring the limits of language modeling," arXiv preprint arXiv:1602.02410, 2016.
[83] G. Melis et al., "On the state of the art of evaluation in neural language models," arXiv preprint arXiv:1707.05589, 2017.
[84] S. Merity et al., "Regularizing and optimizing LSTM language models," arXiv preprint arXiv:1708.02182, 2017.
[85] K. M. Hermann et al., "Teaching machines to read and comprehend," in Advances in neural information processing systems, 2015.
[86] I. Goodfellow et al., "Generative adversarial nets," in Advances in neural information processing systems, 2014.
[87] S. Hochreiter and J. Schmidhuber, "Long short-term memory," in Journal of Neural computation, 1997.
[88] S. Narayan et al., "Don't give me the details, just the summary! topic-aware convolutional neural networks for extreme summarization," arXiv preprint arXiv:1808.08745, 2018.
[89] P. Over and J. Yen, "An introduction to DUC-2004," National Institute of Standards and Technology, 2004.
[90] J. G. Carbonell and J. Goldstein, "The use of MMR, diversity-based reranking for reordering documents and producing summaries," in Special Interest Group on Information Retrieval, 1998.
[91] X. Wan and J. Yang, "Improved affinity graph based multi-document summarization," in North American Chapter of the Association for Computational Linguistics, 2006.
[92] E. Baralis et al., "GRAPHSUM: Discovering correlations among multiple terms for graph-based summarization," in Journal of Information Sciences, 2013.
[93] M. Litvak and M. Last, "Graph-based keyword extraction for single-document summarization," in workshop on Multi-source Multilingual Information Extraction and Summarization, 2008.
[94] T. H. Haveliwala, "Topic-sensitive pagerank," in International conference on World Wide Web, 2002.
[95] T. Polajnar et al., "An exploration of discourse-based sentence spaces for compositional distributional semantics," in Workshop on Linking Computational Models of Lexical, Sentential and Discourse-level Semantics, 2015.
[96] Z. Harris, "Distributional structure," in The Structure of Language, 1954.
[97] V. N. Gudivada and V. V. Raghavan, "Content based image retrieval systems," in IEEE Workshop on Content-Based Access of Image and Video Libraries, 1995.
[98] D. L. Swets and J. J. Weng, "Using discriminant eigenfeatures for image retrieval," in IEEE Transactions on pattern analysis machine intelligence, 1996.
[99] C. Schmid and R. Mohr, "Local grayvalue invariants for image retrieval," in IEEE transactions on pattern analysis machine intelligence, 1997.
[100] Y. Rui et al., "Image retrieval: current techniques, promising directions, and open issues," in Journal of Visual Comm. and Image Representation, 1999.
[101] C. G. Snoek and M. Worring, "Concept-based video retrieval," Foundations Trends® in Information Retrieval, 2009.
[102] H. J. Zhang et al., "An integrated system for content-based video retrieval and browsing," in Journal of Pattern recognition, 1997.
[103] N. Dimitrova et al., "Applications of video-content analysis and retrieval," in Journal of IEEE multimedia, 2002.
[104] Y.-G. Jiang et al., "Towards optimal bag-of-features for object categorization and semantic video retrieval," in ACM Image and video retrieval, 2007.
[105] D. R. Lowery, "Sound storage and sound retrieval system having peripheral with hand operable switches," ed: Google Patents, 1995.
[106] R. F. Lyon et al., "Sound retrieval and ranking using sparse auditory representations," in Journal of Neural computation, 2010.
[107] S. Wake and T. Asahi, "Sound retrieval with intuitive verbal expressions," Georgia Institute of Technology, 1998.
[108] G. Salton and C. Buckley, "Term-weighting approaches in automatic text retrieval," in Journal of Information processing management, 1988.
[109] G. Salton, "Developments in automatic text retrieval," in Journal of Science, 1991.
[110] J. Sivic and A. Zisserman, "Video Google: A text retrieval approach to object matching in videos," in Computer Vision, 2003.
[111] G. Salton et al., "Extended boolean information retrieval," Cornell University, 1982.
[112] O. Chapelle and V. Vapnik, "Model selection for support vector machines," in Advances in neural information processing systems, 2000.
[113] H. P. Luhn, "A statistical approach to mechanized encoding and searching of literary information," in Journal of IBM Research and Development, 1957.
[114] K. Sparck Jones, "A statistical interpretation of term specificity and its application in retrieval," in Journal of documentation, 1972.
[115] T. YuClement, "On the construction of effective vocabularies for information retrieval," in Association for Computing Machinery's Special Interest Group on Information Retrieval, 1973.
[116] G. Salton and C.-S. Yang, "On the specification of term values in automatic indexing," in Journal of documentation, 1973.
[117] E. Efthimiadis, "Query expansion," Annual Review of Information Systems and Technology, 1996.
[118] Y. Qiu and H.-P. Frei, "Concept based query expansion," in Association for Computing Machinery's Special Interest Group on Information Retrieval, 1993.
[119] E. M. Voorhees, "Query expansion using lexical-semantic relations," in Association for Computing Machinery's Special Interest Group on Information Retrieval, 1994.
[120] M. Mitra et al., "Improving automatic query expansion," in Association for Computing Machinery's Special Interest Group on Information Retrieval, 1998.
[121] G. A. Miller, "WordNet: a lexical database for English," in Journal of Communications of the ACM, 1995.
[122] L. Denoyer and P. Gallinari, "The wikipedia xml corpus," in International Workshop of the Initiative for the Evaluation of XML Retrieval, 2006.
[123] G. Salton and C. Buckley, "Improving retrieval performance by relevance feedback," in Journal of American Society for Information Science, 1990.
[124] G. Cao et al., "Selecting good expansion terms for pseudo-relevance feedback," in Association for Computing Machinery's Special Interest Group on Information Retrieval, 2008.
[125] H.-M. Wang et al., "MATBN: A mandarin chinese broadcast news corpus," in Journal of Computational Linguistics & Chinese Language Processing: Special Issue on Annotated Speech Corpora, 2005
[126] D. Povey et al., "The kaldi speech recognition toolkit," in IEEE workshop on automatic speech recognition and understanding, 2011.
[127] R. V. Shannon et al., "Speech recognition with primarily temporal cues," in Journal of Science, 1995.
[128] A. Graves et al., "Speech recognition with deep recurrent neural networks," in IEEE acoustics, speech and signal processing, 2013.
[129] L. Lovász, "Submodular functions and convexity," in Mathematical Programming The State of the Art: Springer, 1983.
[130] C. Lavania and J. Bilmes, "Auto-summarization: a step towards unsupervised learning of a submodular mixture," in SIAM Data Mining, 2019.
[131] J. L. Ba et al., "Layer normalization," arXiv preprint arXiv:1607.06450, 2016.
[132] E. Sandhaus, "The new york times annotated corpus," in Linguistic Data Consortium, Philadelphia, 2008.
[133] G. Klein et al., "Opennmt: open-source toolkit for neural machine translation," arXiv preprint arXiv:1701.02810, 2017.

全文公開日期 2025/02/11 (校內網路)
全文公開日期 2025/02/11 (校外網路)
全文公開日期 2025/02/11 (國家圖書館：臺灣博碩士論文系統)

簡易檢索 / 詳目顯示

相關論文