文本摘要研究进展与趋势

doi:10.11959/j.issn.2096-109x.2018048

摘要/Abstract

摘要：

随着互联网上的信息呈爆炸式增长，如何从海量信息中提取有用信息成了一个关键的技术问题。文本摘要技术能够从大数据中压缩提炼出精炼简洁的文档信息，有效降低用户的信息过载问题，成为研究热点。分类整理分析了近些年来国内外的文本摘要方法及其具体实现，将传统方法和深度学习摘要方法的优缺点进行了对比分析，并对今后的研究方向进行了合理展望。

关键词: 大数据, 文本摘要, 机器学习, 传统方法, 深度学习

Abstract:

With the explosive growth of information on the Internet,how to extract useful information from massive information has become a key technical issue.The text summarization technology can compress and extract refined and concise document information from big data,effectively reducing the user information overload problem,and it has become a research hotspot.The domestic and foreign text summarization methods and their concrete realization in recent years were analyzed,the advantages and disadvantages between traditional methods and deep learning summary methods were compared,and a reasonable outlook for future research directions was made.

Key words: big data, text summarization, machine learning, traditional methods, deep learning

中图分类号:

TP393

明拓思宇, 陈鸿昶. 文本摘要研究进展与趋势[J]. 网络与信息安全学报, 2018, 4(6): 1-10.

Tuosiyu MING, Hongchang CHEN. Research progress and trend of text summarization[J]. Chinese Journal of Network and Information Security, 2018, 4(6): 1-10.

图/表 1

表1

各文本摘要方法的优缺点"

方法	优点	缺点
基于统计学方法	依据文本形式上的规律，简单直观，避免考虑复杂的句法、语法结构，易于实现且应用广泛，无需训练数据，执行速度快	只是单纯利用了单词表层特征，没有充分挖掘词义关系和语义特征，存在较大局限性
基于外部语义资源方法	在统计学方法的基础上利用词间关系、词义关系进行了改进，使文本摘要的语义性能得到了一定的提高	受收录词汇的限制比较大；对于文章题目依赖程度较高；分词对关键词的影响较大；相似度阈值的选取对构建词汇链有影响。语法语义结构不连贯
基于图排序方法	适用于结构较为松散且涉及主题较多的结构；计算句子权重的同时可以充分考虑词汇之间、词组之间或句子之间的全局关系；无监督，语言独立，不需要对大量语料进行处理	通常只考虑了句子节点间的相似性关系，而忽略了文档篇章结构以及句子上下文的信息；相似度计算的好坏决定了关键词和句子重要性排序的正确与否；对数据的利用不够充分；没有考虑信息冗余
基于统计机器学习方法	特征选择和训练分类器的选择上有较大的可供选择范围，还可以综合一些开放性特征提高分类的精度	需要人工标注的数据集；效果严格依赖于训练数据质量的好坏；监督或半监督，执行速度较无监督的方法慢
基于深度学习方法	降低了对人工的依赖，可以高效地进行训练；可以与多种神经网络结构和Sequence-to-Sequence模型结合，生成文本摘要的可读性和准确度高	可解释性差；需要大量人工标注的数据集；由于有复杂的神经网络结构的引入，执行速度慢，需要花费相对较长的时间；对计算机性能有一定的要求

表1

参考文献 69

[1]	CHENG J , LAPATA M . Neural summarization by extracting sentences and words[J]. ar Xiv preprint ar Xiv:1603.07252, 2016.
[2]	NEMA P , KHAPRA M , LAHA A ,et al. Diversity driven attention model for query-based abstractive summarization[J]. ar Xiv preprint ar Xiv:1704.08300, 2017.
[3]	LI P , LAM W , BING L ,et al. Deep recurrent generative decoder for abstractive text summarization[J]. ar Xiv preprint ar Xiv:1708.00625, 2017.
[4]	BING L , LI P , LIAO Y ,et al. Abstractive multi-document summarization via phrase selection and merging[J]. ar Xiv preprint ar Xiv:1506.01597, 2015.
[5]	LI C , QIAN X , LIU Y . Using supervised bigram-based ilp for extractive summarization[C]// The 51st Annual Meeting of the Association for Computational Linguistics. 2013: 1004-1013.
[6]	VEENA G , GUPTA D , JAGANADH J ,et al. A graph based conceptual mining model for abstractive text summarization[J]. Indian Journal of Science and Technology, 2016,9(S1).
[7]	DANESH S , SUMNER T , MARTIN J H . Sgrank:combining statistical and graphical methods to improve the state of the art in unsupervised keyphrase extraction[C]// The Fourth Joint Conference on Lexical and Computational Semantics. 2015: 117-126.
[8]	FLORESCU C , CARAGEA C . Position Rank:an unsupervised approach to keyphrase extraction from scholarly documents[C]// The 55th Annual Meeting of the Association for Computational Linguistics. 2017: 1105-1115.
[9]	LUHN H P . The automatic creation of literature abstracts[J]. IBM Journal of Research and Development, 1958,2(2): 159-165.
[10]	BAXENDALE P B . Machine-made index for technical literature—an experiment[J]. IBM Journal of Research and Development, 1958,2(4): 354-361.
[11]	EDMUNDSON H P . New methods in automatic extracting[J]. Journal of the ACM, 1969,16(2): 264-285.
[12]	SALTON G , YU C T . On the construction of effective vocabularies for information retrieval[C]// ACM SIGIR Forum. 1973: 48-60.
[13]	施聪莺, 徐朝军, 杨晓江 . TFIDF 算法研究综述[J]. 计算机应用, 2009,29(B06): 167-170.
	SHI C Y , XU C J , YANG X J . Study of TFIDF algorithm[J]. Journal of Computer Applications, 2009,29(B06): 167-170.
[14]	徐文海, 温有奎 . 一种基于 TFIDF 方法的中文关键词抽取算法[J]. 情报理论与实践, 2008,31(2): 298-302.
	XU W H , WEN Y K . A Chinese keyword extraction algorithm based on TFIDF method[J]. Information Studies:Theory ＆ Application, 2008,31(2): 298-302.
[15]	SUQIN Z B S H M . An improved text feature weighting algorithm based on TFIDF[J]. Computer Applications and Software, 2011,2: 7.
[16]	李静月, 李培峰, 朱巧明 . 一种改进的 TFIDF 网页关键词提取方法[J]. 计算机应用与软件, 2011,28(5): 25-27.
	LI J Y , LI P F , ZHU Q M . An improved TFIDF-based approach to extract key words from Wed pages[J]. Computer Applications and Software, 2011,28(5): 25-27.
[17]	EL-BELTAGY S R , RAFEA A . Kp-miner:participation in semeval-2[C]// The 5th International Workshop on Semantic Evaluation, 2010: 190-193.
[18]	DANESH S , SUMNER T , MARTIN J H . Sgrank:combining statistical and graphical methods to improve the state of the art in unsupervised keyphrase extraction[C]// The Fourth Joint Conference on Lexical and Computational Semantics. 2015: 117-126.
[19]	FLORESCU C , CARAGEA C . Position Rank:an unsupervised approach to keyphrase extraction from scholarly documents[C]// The 55th Annual Meeting of the Association for Computational Linguistics, 2017: 1105-1115.
[20]	PADMALAHARI E , KUMAR D V N S , PRASAD S . Automatic text summarization with statistical and linguistic features using successive thresholds[C]// 2014 International Conference on Advanced Communication Control and Computing Technologies (ICACCCT). 2014: 1519-1524.
[21]	MILLER G A . Word Net:a lexical database for English[J]. Communications of the ACM, 1995,38(11): 39-41.
[22]	BARZILAY R , ELHADAD M . Using lexical chains for text summarization[J]. Advances in Automatic Text Summarization, 1999: 111-121.
[23]	JAIN A , GAUR A . Summarizing long historical documents using significance and utility calculation using Word Net[J]. Imperial Journal of Interdisciplinary Research, 2017,3(3).
[24]	SILBER H G , MCCOY K F . Efficient text summarization using lexical chains[C]// The 5th International Conference on Intelligent user interfaces. ACM, 2000: 252-255.
[25]	KOLLA M . Automatic text summarization using lexical chains:algorithms and experiments[D]. University of Lethbridge, 2004.
[26]	POURVALI M , ABADEH M S . Automated text summarization base on lexicales chain and graph using of wordnet and wikipedia knowledge base[J]. ar Xiv preprint ar Xiv:1203.3586, 2012.
[27]	HOU S , HUANG Y , FEI C ,et al. Holographic lexical chain and its application in chinese text summarization[C]// Asia-Pacific Web (APWeb) and Web-Age Information Management (WAIM) Joint Conference on Web and Big Data. 2017: 266-281.
[28]	LYNN H M , CHOI C , KIM P . An improved method of automatic text summarization for Web contents using lexical chain with semantic-related terms[J]. Soft Computing, 2018,22(12): 4013-4023.
[29]	PAL A R , SAHA D . An approach to automatic text summarization using Word Net[C]// IEEE International Conference on Advance Computing Conference (IACC). 2014: 1169-1173.
[30]	PAGE L , BRIN S , MOTWANI R ,et al. The Page Rank citation ranking:bringing order to the Web[R]. Stanford Info Lab, 1999.
[31]	KLEINBERG J M , KUMAR R , RAGHAVAN P ,et al. The Web as a graph:measurements,models,and methods[C]// International Computing and Combinatorics Conference, 1999: 1-17.
[32]	MIHALCEA R , . Graph-based ranking algorithms for sentence extraction,applied to text summarization[C]// Proceedings of the ACL 2004 on Interactive Poster And Demonstration Sessions. Association for Computational Linguistics, 2004:20.
[33]	WAN X , XIAO J . Single document keyphrase extraction using neighborhood knowledge[C]// AAAI. 2008,8: 855-860.
[34]	GOLLAPALLI S D , CARAGEA C . Extracting keyphrases from research papers using citation networks[C]// AAAI. 2014: 1629-1635.
[35]	KHAN A , SALIM N , FARMAN H ,et al. Abstractive text summarization based on improved semantic graph approach[J]. International Journal of Parallel Programming, 2018: 1-25.
[36]	AL-KHASSAWNEH Y A , SALIM N , JARRAH M . Improving triangle-graph based text summarization using hybrid similarity function[J]. Indian Journal of Science and Technology, 2017,10(8).
[37]	WEI F , LI W , LU Q ,et al. A document-sensitive graph model for multi-document summarization[J]. Knowledge and Information Systems, 2010,22(2): 245-259.
[38]	GE S S , ZHANG Z , HE H . Weighted graph model based sentence clustering and ranking for document summarization[C]// 2011 4th International Conference on Interaction Sciences (ICIS). 2011: 90-95.
[39]	NGUYEN-HOANG T A , NGUYEN K , TRAN Q V . TSGVi:a graph-based summarization system for Vietnamese documents[J]. Journal of Ambient Intelligence and Humanized Computing, 2012,3(4): 305-313.
[40]	耿焕同, 蔡庆生, 赵鹏 ,等. 一种基于词共现图的文档自动摘要研究[J]. 情报学报, 2005,24(6): 652-1.
	GENG H T , CAI Q S , ZHAO P ,et al. Research on document automatic summarization based on word co-occurrence[J]. Journal of the China Society for Scientific and Technical Information, 2005,24(6): 652-1.
[41]	SEHGAL S , KUMAR B , RAMPAL L ,et al. A modification to graph based approach for extraction based automatic text summarization[M]// Progress in Advanced Computing and Intelligent Engineering. Singapore Springer Press, 2018: 373-378.
[42]	YOUSEFI-AZAR M , HAMEY L . Text summarization using unsupervised deep learning[J]. Expert Systems with Applications, 2017,68: 93-105.
[43]	ARRAS L , HORN F , MONTAVON G ,et al. What is relevant in a text document? an interpretable machine learning approach[J]. PloSone, 2017,12(8):e0181142
[44]	THU H N T . An optimization text summarization method based on naive bayes and topic word for single syllable language[J]. Applied Mathematical Sciences, 2014,8(3): 99-115.
[45]	SILVA G , FERREIRA R , LINS R D ,et al. Automatic text document summarization based on machine learning[C]// 2015 ACM Symposium on Document Engineering. ACM, 2015: 191-194.
[46]	NISHIKAWA H , ARITA K , TANAKA K ,et al. Learning to generate coherent summary with discriminative hidden semi-markov model[C]// The 25th International Conference on Computational Linguistics:Technical Papers. 2014: 1648-1659.
[47]	ALLAHYARI M , POURIYEH S , ASSEFI M ,et al. A brief survey of text mining:classification,clustering and extraction techniques[J]. ar Xiv preprint ar Xiv:1707.02919, 2017.
[48]	KUPIEC J , PEDERSEN J , CHEN F . A trainable document summarizer[C]// The 18th annual International ACM SIGIR Conference on Research and Development in Information Retrieval. 1995: 68-73.
[49]	CONROY J M,O'LEARY D P , . Text summarization via hidden markov models[C]// The 24th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. 2001: 406-407.
[50]	LIN C Y , . Training a selection function for extraction[C]// The Eighth International Conference on Information and Knowledge Management. ACM, 1999: 55-62.
[51]	HINTON G E , OSINDERO S , TEH Y W . A fast learning algorithm for deep belief nets[J]. Neural Computation, 2006,18(7): 1527-1554.
[52]	MRK?I? N , VULI? I , SéAGHDHA D ó ,et al. Semantic specialisation of distributional word vector spaces using monolingual and cross-lingual constraints[J]. ar Xiv preprint ar Xiv:1706.00374, 2017.
[53]	XIONG Z , SHEN Q , WANG Y ,et al. Paragraph vector representation based on word to vector and CNN learning[J]. CMC:Computers,Materials ＆ Continua, 2018,55(2): 213-227.
[54]	WANG X , ZHANG H , LIU Y . Sentence vector model based on implicit word vector expression[J]. IEEE Access, 2018,6: 17455-17463.
[55]	SUTSKEVER I , VINYALS O , Le Q V . Sequence to sequence learning with neural networks[C]// Advances in neural information processing systems. 2014: 3104-3112.
[56]	NALLAPATI R , XIANG B , ZHOU B . Sequence-to-sequence rnns for text summarization[J]. ar Xiv preprint ar Xiv:1602.06023v1, 2016.
[57]	RUSH A M , CHOPRA S , WESTON J . A neural attention model for abstractive sentence summarization[J]. ar Xiv preprint ar Xiv:1509.00685, 2015.
[58]	CHOPRA S , AULI M , RUSH A M . Abstractive sentence summarization with attentive recurrent neural networks[C]// The 2016 Conference of the North American Chapter of the Association for Computational Linguistics:Human Language Technologies. 2016: 93-98.
[59]	NALLAPATI R , ZHOU B , GULCEHRE C ,et al. Abstractive text summarization using sequence-to-sequence RNNS and beyond[J]. ar Xiv preprint ar Xiv:1602.06023v5, 2016.
[60]	CAO Z , LI W , LI S ,et al. Attsum:Joint learning of focusing and summarization with neural attention[J]. ar Xiv preprint ar Xiv:1604.00125, 2016.
[61]	SEE A , LIU P J , Manning C D . Get to the point:summarization with pointer-generator networks[J]. ar Xiv preprint ar Xiv:1704.04368, 2017.
[62]	ABADI M , BARHAM P , CHEN J ,et al. Tensor Flow:a system for large-scale machine learning[C]// OSDI. 2016: 265-283.
[63]	SUTSKEVER I , VINYALS O , LE Q V . Sequence to sequence learning with neural networks[C]// Advances in Neural Information Processing Systems. 2014: 3104-3112.
[64]	GEHRING J , AULI M , GRANGIER D ,et al. Convolutional sequence to sequence learning[J]. ar Xiv preprint ar Xiv:1705.03122, 2017.
[65]	LIU L , LU Y , YANG M ,et al. Generative adversarial network for abstractive text summarization[J]. ar Xiv preprint ar Xiv:1711.09357, 2017.
[66]	GOODFELLOW I , POUGET-ABADIE J , MIRZA M ,et al. Generative adversarial nets[C]// Advances in Neural Information Processing Systems. 2014: 2672-2680.
[67]	TAN J , WAN X , XIAO J . Abstractive document summarization with a graph-based attentional neural model[C]// The 55th Annual Meeting of the Association for Computational Linguistics. 2017: 1171-1181.
[68]	MIHALCEA R , TARAU P . Textrank:bringing order into text[C]// Proceedings of the 2004 Conference on Empirical Methods in Natural Language Processing. 2004.
[69]	ERKAN G , RADEV D R . Lexrank:graph-based lexical centrality as salience in text summarization[J]. Journal of Artificial Intelligence Research, 2004,22: 457-479.