|
|
|
| Research on Abstract Structure Function Automatic Recognition Based on Full Character Semantics |
| Shen Si1, Hu Haotian2, Ye Wenhao2, Wang Dongbo2 |
1. School of Economics & Management, Nanjing University of Science and Technology, Nanjing 210094;
2. College of Information Science and Technology, Nanjing Agricultural University, Nanjing 210095 |
|
|
|
|
Abstract The structure of each academic-literature abstract has a specific function. However, there are relatively few studies on the automatic recognition of the structural abilities of academic abstracts at present; furthermore, these studies have some problems, such as methods that are too traditional, as well as insignificant recognition. Based on the deep learning method of the LSTM-CRF model with sequence properties, this paper constructed an automatic structure recognition model that uses the semantic information contained in all characters in the abstract, and compared the result with SVM models without sequence properties, RNN, CRF and LSTM with sequence properties in multiple angles by taking the character as the basic semantic unit. The model proposed in this paper achieved remarkable results in accuracy, recall, and F-value in structure recognition, with the highest F-value reaching 85.47%. Compared with the models of RNN, LSTM, CRF, and SVM, its performance is enhanced by 33.63%, 32.81%, 39.13%, and 38.33%, respectively.
|
|
Received: 15 December 2017
|
|
|
|
［1］ 黄玲, 姜小兰. 我国科技期刊摘要存在的问题及原因浅析[J]. 长江科学院院报, 2014, 31(7): 114-118.
［2］ 张郁佳. 中英语言学学术期刊摘要的元功能分析[J]. 黑龙江生态工程职业学院学报, 2017, 30(4): 148-150.
［3］ 高建群, 吴玲, 施业. 学术论文摘要的规范表达[J]. 东南大学学报(哲学社会科学版), 2003(2): 114-117.
［4］ Silva J, Coheur L, Mendes A C, et al. From symbolic to sub-symbolic information in question classification[J]. Artificial Intelligence Review, 2011, 35(2): 137-154.
［5］ Cui L, Zhang D, Liu S, et al. Learning topic representation for SMT with neural networks[C]//Proceedings of Conference of the Association for Computational Linguistics. Stroudsburg: Association for Computational Linguistics, 2014: 133-143.
［6］ Wang D, Nyberg E. A long short-term memory model for answer sentence selection in question answering[C]//Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing. Stroudsburg: Association for Computational Linguistics, 2015: 707-712.
［7］ 郑秋生, 翟琳琳. 基于改进Rocchio算法的短文本自动分类研究[J]. 中原工学院学报, 2013, 24(1): 70-73.
［8］ 林小俊, 张猛, 暴筱, 等. 基于概念网络的短文本分类方法[J]. 计算机工程, 2010, 36(21): 4-6.
［9］ 孙建旺, 吕学强, 张雷瀚. 基于语义与最大匹配度的短文本分类研究[J]. 计算机工程与设计, 2013, 34(10): 3613-3618.
［10］ 张群, 王红军, 王伦文. 词向量与LDA相融合的短文本分类方法[J]. 现代图书情报技术, 2016(12): 27-35.
［11］ 李湘东, 曹环, 丁丛, 等. 利用《知网》和领域关键词集扩展方法的短文本分类研究[J]. 现代图书情报技术, 2015(2): 31-38.
［12］ Zhou C, Sun C, Liu Z, et al. A c-LSTM neural network for text classification[J]. arXiv Preprint arXiv: 1511.08630, 2015.
［13］ Dong L, Wei F, Zhou M, et al. Question answering over freebase with multi-column convolutional neural networks[C]//Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing. Stroudsburg: Association for Computational Linguistics, 2015: 260-269.
［14］ Qu Y, Liu J, Kang L, et al. Question answering over freebase via attentive RNN with similarity matrix based CNN[J]. arXiv: 1804.03317, 2018.
［15］ Zhang D, Wang D. Relation classification via recurrent neural network[J]. arXiv Preprint arXiv: 1508.01006, 2015.
［16］ Ji W. Structure regularized bidirectional recurrent convolutional neural network for relation classification[J]. arXiv Preprint arXiv: 1711.02509, 2017.
［17］ 孙晓, 彭晓琪, 胡敏, 等. 基于多维扩展特征与深度学习的微博短文本情感分析[J]. 电子与信息学报, 2017, 39(9): 2048-2055.
［18］ 殷亚博, 杨文忠, 杨慧婷, 等. 基于卷积神经网络和KNN的短文本分类算法研究[J]. 计算机工程, 2018, 44(7): 193-198.
［19］ 彭玉青, 宋初柏, 闫倩, 等. 基于VDCNN和LSTM混合模型的中文文本分类研究[J]. 计算机工程, 2017, 43(11): 123-130.
［20］ 薛涛, 王雅玲, 穆楠. 基于词义消歧的卷积神经网络文本分类模型[J]. 计算机应用研究, 2018, 35(10): 1-8.
［21］ 陆伟, 黄永, 程齐凯. 学术文本的结构功能识别——功能框架及基于章节标题的识别[J]. 情报学报, 2014, 33(9): 979-985.
［22］ 黄永, 陆伟, 程齐凯. 学术文本的结构功能识别——基于章节内容的识别[J]. 情报学报, 2016, 35(3): 293-300.
［23］ 黄永, 陆伟, 程齐凯, 等. 学术文本的结构功能识别——基于段落的识别[J]. 情报学报, 2016, 35(5): 530-538.
［24］ 王立非, 刘霞. 英语学术论文摘要语步结构自动识别模型的构建[J]. 外语电化教学, 2017(2): 45-50, 64.
［25］ Pham T H, Le-Hong P. End-to-end recurrent neural network models for vietnamese named entity recognition: Word-level vs. character-level[C]// Proceedings of the 15th International Conference of the Pacific Association for Computational Linguistics. Springer, 2017: 46-53.
［26］ Hochreiter S, Schmidhube R J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
［27］ Cross J, Huang L. Incremental parsing with minimal features using bi-directional LSTM[J]. arXiv Preprint arXiv: 1606.06406, 2016.
［28］ Hearst M A, Dumais S T, Osuna E, et al. Support vector machines[J]. IEEE Intelligent Systems and Their applications, 1998, 13(4): 18-28.
［29］ Lafferty J, McCallum A, Pereira F. Conditional random fields: Probabilistic models for segmenting and labeling sequence data[C]// Proceedings of the Eighteenth International Conference on Machine Learning, 2001: 282-289.
［30］ Huang Z, Xu W, Yu K. Bidirectional LSTM-CRF models for sequence tagging[J]. arXiv Preprint arXiv: 1508.01991, 2015. |
|
|
|
