构建基于科技文献知识的人工智能引擎

doi:10.13998/j.cnki.issn1002-1248.20-0797

农业图书情报学报 ›› 2021, Vol. 33 ›› Issue (1): 17-31.doi: 10.13998/j.cnki.issn1002-1248.20-0797

所属专题：智慧服务

构建基于科技文献知识的人工智能引擎

张智雄^1,2,3,4, 刘欢^1,2,4, 于改红¹

1.中国科学院文献情报中心,北京 100190;
2.中国科学院大学经济与管理学院图书情报与档案管理系,北京 100190;
3.中国科学院武汉文献情报中心,武汉 430071;
4.科技大数据湖北省重点实验室,武汉 430071

收稿日期:2020-09-29 出版日期:2021-01-05 发布日期:2021-02-05
作者简介:张智雄（ORCID：0000-0003-1596-7487）,男,博士,二级研究员,博导,研究方向为图书情报学智能信息处理。刘欢（1995- ）,男,博士生,研究方向为深度学习、文本挖掘。于改红（ORCID：0000-0003-1301-2871）,女,博士,馆员,研究方向为文本挖掘、情报分析
基金资助:
国家科技图书文献中心（NSTL）“下一代开放知识服务平台总体设计及关键技术研发”专项“基于深度学习的科技论文语步标注技术工具研发”（2019XM55）; 中国科学院文献情报能力建设专项课题“基于科技文献知识的人工智能（AI）引擎建设”

Building an Artificial Intelligence Engine Based on Scientific and Technological Literature Knowledge

ZHANG Zhixiong^1,2,3,4, LIU Huan^1,2,4, YU Gaihong¹

1. National Science Library, Chinese Academy of Sciences, Beijing 100190;
2. University of Chinese academy of sciences, Beijing 100049;
3. Wuhan Library, Chinese Academy of Sciences, Wuhan 430071;
4. Hubei Key Laboratory of Big Data in Science and Technology, Wuhan 430071

Received:2020-09-29 Online:2021-01-05 Published:2021-02-05

摘要/Abstract

摘要： [目的/意义]如何利用科技文献中的知识对深度学习算法模型进行训练改进,以获取知识、发现知识是情报研究的重要课题。为充分挖掘和利用科技文献中的知识,本文提出构建基于科技文献知识的人工智能引擎的目标。[方法/过程]本文以文献情报工作为出发点,立足科技文献是人类知识的最重要载体,探究人工智能取得飞速突破的本质所在,创新性提出了文献情报领域从“科技文献库”转变为“科技知识引擎”的建设思路。[结果/结论]本文论述了中国科学院文献情报中心在构建基于科技文献知识的人工智能（AI）引擎的建设实践,探讨了利用深度学习技术挖掘知识以服务情报研究的方法,以期为学界提供参考。

关键词: 科技文献, 人工智能知识引擎, 预训练语言模型, 微调模型, AI引擎构建实践

Abstract: [Purpose/Significance] How to use the knowledge in the scientific and technological literature to train and improve the model of deep learning algorithm, and acquire knowledge and discover knowledge is an important subject of information research. In order to fully mine and utilize the value of literature knowledge, this paper proposes the goal of building an artificial intelligence (AI) engine based on scientific and technological literature knowledge. [Method/Process] It chooses the literature and information science work as the starting point and takes the scientific and technological literature as the most important carrier of human knowledge. This paper explores the essence of the rapid breakthrough of AI, and innovatively puts forward the construction idea of "science and technology knowledge engine" which is the transformation from "science and technology literature library" in the field of information science. [Results/Conclusions] This paper discusses the construction practice of AI engine based on scientific and technological literature knowledge and explores the method of using the deep learning technology to excavate knowledge to serve information research, so as to provide reference for peers.

Key words: scientific and technological literature, artificial intelligence (AI) knowledge engine, pre-trained language model, fine-tunning model, construction practice of AI engine

中图分类号:

G250

张智雄, 刘欢, 于改红. 构建基于科技文献知识的人工智能引擎[J]. 农业图书情报学报, 2021, 33(1): 17-31.

ZHANG Zhixiong, LIU Huan, YU Gaihong. Building an Artificial Intelligence Engine Based on Scientific and Technological Literature Knowledge[J]. Journal of Library and Information Science in Agriculture, 2021, 33(1): 17-31.

参考文献

[1] BELTAGY I, KYLE L, ARMAN C.SciBERT: A pretrained language model for scientific text[J]. arXiv preprint arXiv:1903.10676, 2019.
[2] WAN H, ZHANG Y, ZHANG J, et al.AMiner: Search and mining of academic social networks[J]. Data intelligence, 2019, 1(1): 58-76.
[3] SHOHAM Y, PERRAULT R, BRYNJOLFSSON E, et al.Artificial intelligence index 2017 annual report[J]. Artificial intelligence index, 2017.
[4] WU Y, SCHUSTER M, CHEN Z, et al. Google's neural machine translation system: Bridging the gap between human and machine translation[J]. ArXiv preprint arxiv:1609.08144, 2016.
[5] MCCULLOCH W S, PITTS W.A logical calculus of the ideas immanent in nervous activity[J]. The bulletin of mathematical biophysics, 1943, 5(4): 115-133.
[6] HOCHREITER S, SCHMIDHUBER J.Long short-term memory[J]. Neural computation, 1997, 9(8): 1735-1780.
[7] LECUN Y, BOSER B, DENKER J S, et al.Backpropagation applied to handwritten zip code recognition[J]. Neural computation, 1989, 1(4): 541-551.
[8] GOODFELLOW I, BENGIO Y, COURVILLE A.Deep learning[M]. Cambridge: MIT press, 2016.
[9] 雷明. 机器学习与应用[M]. 北京: 清华大学出版社, 2019.
LEI M.Machine learning and application[M]. Beijing: Tsinghua university press, 2019.
[10] DEVLIN J, CHANG M W, LEE K, et al. Bert: Pre-training of deep bidirectional transformers for language understanding[J]. ArXiv preprint arxiv:1810.04805, 2018.
[11] PETERS M E, NEUMANN M, IYYER M, et al. Deep contextualized word representations[J]. ArXiv preprint arxiv:1802.05365, 2018.
[12] RADFORD A, NARASIMHAN K, SALIMANS T, et al.Improving language understanding by generative pre-training[J]. URL https://s3-us-west-2.amazonaws.com/openai-assets/researchcovers/languageunsupervised/language understanding paper.pdf, 2018.
[13] RADFORD A, WU J, CHILD R, et al.Language models are unsupervised multitask learners[J]. OpenAI blog, 2019, 1(8).
[14] YANG Z, DAI Z, YANG Y, et al.XLNet: Generalized autoregressive pretraining for language understanding[J]. ArXiv preprint arxiv:1906.08237, 2019.
[15] LIU Y, OTT M, GOYAL N, et al.Roberta: A robustly optimized Bert pretraining approach[J]. ArXiv preprint arxiv:1907.11692, 2019.
[16] WANG A, SINGH A, MICHAEL J, et al. Glue: A multi-task benchmark and analysis platform for natural language understanding[J]. ArXiv preprint arxiv:1804.07461, 2018.
[17] SUN Y, WANG S, LI Y, et al.ERNIE: Enhanced representation through knowledge integration[J]. ArXiv preprint arxiv:1904.09223,2019.
[18] SUN Y, WANG S, LI Y, et al.ERNIE 2.0: a continual pre-training framework for language understanding[J]. ArXiv preprint arxiv:1907.12412, 2019.
[19] CUI Y, CHE W, LIU T, et al.Pre-Training with whole word masking for Chinese BERT[J]. ArXiv preprint arxiv:1906.08101, 2019.
[20] OpenCLaP. 多领域开源中文预训练语言模型仓库项目简介[EB/OL]. [2020-08-07]. http://zoo.thunlp.org/.
OpenCLaP. Open Chinese language pre-trained model zoo project brief[EB/OL]. [2020-08-07]. http://zoo.thunlp.org/.
[21] 谢玮, 沈一, 马永征. 基于图计算的论文审稿自动推荐系统[J]. 计算机应用研究, 2016, 33(3): 798-801.
XIE W, SHEN Y, MA Y Z.Recommendation system for paper reviewing based on graph computing[J]. Application research of computers, 2016, 33(3): 798-801.
[22] WANG D, LIANG Y, XU D, et al.A Content-Based recommender system for computer science publications[J]. Knowledge-Based systems, 2018: S0950705118302107.
[23] 于改红, 张智雄, 马娜. 科技文献语篇元素自动标注模型研究综述[J]. 图书情报工作, 2018, 62(15): 132-144.
YU G H, ZHANG Z X, MA N.Overview of science and technology literature discourse elements automatic annotation model research[J]. Library and information service, 2018, 62(15): 132-144.
[24] HéLèNE D R, FALQUET G. An automated annotation process for the SciDocAnnot scientific document model[C]//5th international workshop on semantic digital archives, 2015.
[25] LIAKATA M, TEUFEL S, SIDDHARTHAN A, et al.Corpora for the conceptualisation and zoning of scientific papers[C]//Proceedings of the international conference on language resources and evaluation, LREC 2010, 17-23 may 2010, Valletta, Malta: DBLP, 2010.
[26] TEUFEL S, SIDDHARTHAN A, BATCHELOR C.Towards discipline-independent argumentative zoning[C]//The 2009 conference. association for computational linguistics, 2009.
[27] FISAS B, RONZANO F, SAGGION H.A multi-layered annotated corpus of scientific papers[C]//LREC 2016, 2016.
[28] 张智雄, 刘欢, 丁良萍, 等. 不同深度学习模型的科技论文摘要语步识别效果对比研究[J]. 数据分析与知识发现, 2019, 3(12):1-9.
ZHANG Z X, LIU H, DING L P, et al.Identifying moves of research abstracts with deep learning methods[J]. Data analysis and knowledge discovery, 2019, 3(12): 1-9.
[29] YU G, ZHANG Z, LIU H, et al.Masked sentence model based on BERT for move recognition in medical scientific abstracts[J], 2019.
[30] DERNONCOURT F, LEE J Y. Pubmed 200k rct: a dataset for sequential sentence classification in medical abstracts[J]. ArXiv preprint arxiv:1710.06071, 2017.
[31] JIN D, PETER S. Hierarchical neural networks for sequential sentence classification in medical scientific abstracts[J]. arXiv preprint arXiv:1808.06161, 2018.
[32] 赵旸, 张智雄, 刘欢, 等. 基于BERT模型的中文医学文献分类研究[J/OL]. 数据分析与知识发现: 1-12[2020-08-05]. http://kns.cnki.net/kcms/detail/10.1478.G2.20200603.1457.004.html.
ZHAO Y, ZHANG Z X, LIU H, et al. Classification of Chinese medical literature with BERT model[J/OL]. Data analysis and knowledge discovery: 1-12[2020-08-05]. http://kns.cnki.net/kcms/detail/10.1478.G2.20200603.1457.004.html.
[33] DING L P, ZHANG Z X, LIU H, et al.Automatic keyphrase extraction from scientific Chinese medical abstracts based on character-level sequence labeling. In Wuhan’20: Joint conference on digital library, Wuhan, 2020.
[34] GERARD S, YANG C S, CLEMENT T Y.A theory of term importance in automatic text analysis[J]. Journal of the American society for Info rmation Science, 1975, 26(1): 33-44.
[35] Huang Z H, XU W, YU K. Bidirectional LSTM-CRF models for sequence tagging[J]. arXiv preprint arXiv:1508.01991, 2015.
[36] NAVIGLI, ROBERTO, PAOLA V.Learning word-class lattices for definition and hypernym extraction[C]. Proceedings of the 48th annual meeting of the association for computational linguistics, 2010.
[37] DONG L, YANG N, WANG W, et al.Unified language model pre-training for natural language understanding and generation[C]//Advances in neural information processing systems, 2019: 13063-13075.

构建基于科技文献知识的人工智能引擎

Building an Artificial Intelligence Engine Based on Scientific and Technological Literature Knowledge

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 4

Metrics

本文评价

推荐阅读 0

[1]	常志军, 许丽媛, 于倩倩, 张建勇, 王永吉. 基于生命周期模型的科技文献数据管理体系研究[J]. 农业图书情报学报, 2022, 34(6): 36-49.
[2]	柴苗岭, 邹弈星, 谭荣志, 曾怡, 任运月. 面向农业产业知识服务的科学数据与科技文献关联研究与实践[J]. 农业图书情报学报, 2022, 34(3): 37-50.
[3]	文雯, 刘斌, 赵金国. 科技文献平台资源共享架构与运行机制研究——以贵州省产业创新科技文献服务平台为例[J]. 农业图书情报学报, 2021, 33(3): 68-77.
[4]	王颖. 科技文献内容语义描述模型研究[J]. 农业图书情报学报, 2020, 32(8): 12-24.