科技文献作为科研人员记录科学真理验证过程、实验观测、研究结论等知识情报的载体，是人类科技发展的智慧结晶。在信息爆炸的互联网时代，信息资源的堆积并不等同于知识量的累积，井喷式增长的科技文献令知识成了更稀缺的资源。问题和方法作为科学研究的本质，科研用户会花费大量时间精力用于定位和整合科技文献的该核心知识点，“广而不专”的粗粒度知识服务已经无法满足用户的精准性和快捷性需求。因此，如何深入科技文献的核心内容，集成知识的“碎片化”和系统性，对知识进行由面到点、从点带面的立体组织，成为亟待解决的问题。鉴于此，本文以知识元理论为指导，提出科技文献的细粒度“问题-方法”知识元模型，在此基础上，研究科技文献“问题-方法”知识元的抽取与组织方法，为缓解用户的细粒度知识需求与粗放的知识服务之间的矛盾提供方法基础。

本研究的工作具体包含“问题-方法”知识元模型构建、知识元抽取、多维知识组织和实验验证四个内容。（1）构建科技文献的细粒度“问题-方法”知识元模型。围绕科技文献的研究问题和研究方法，提取代表性的核心内容要素和辅助内容要素，分别从文献资源、内容要素、语义关联三个维度描述科技文献的外部表现和内部特征，构建“问题-方法”知识元语义描述模型，实现对科技文献研究内容的系统描述。（2）提出一种面向“问题-方法”知识元模型的科技文献知识抽取方法。根据内容语句的句式结构和语法成分，从内容要素抽取和解决关系抽取两个方向设计知识元抽取方法，前者在特征句式中引入标志词，增强描述规则在内容要素抽取中的定位能力，后者基于依存句法分析构建内容语句的内容事件链，以语义连续性为准则抽取“问题-方法”对，实现知识元内容与关系的精准抽取。（3）提出一种基于“问题-方法”知识元模型的科技文献细粒度组织方法。根据词汇的三大基础特征——语用、语法、语义，语用层从共有和特有两个角度比较辅助内容要素对，语法关联通过构建权重分配规则计算内容要素对的内部语法结构，语义关联的重点落于内容要素自身的概念贡献度分析和多级关联内容要素集的语义范围拓展，多角度地组织细粒度知识元。（4）基于“问题-方法”知识元模型进行科技文献知识组织实验。利用内容要素描述规则和解决关系抽取算法，抽取科技文献“问题-方法”知识元，并计算各类内容要素间的相似度，结果证明了研究提出的知识元抽取方法的准确性，组织方法的多维性和系统性。

基于“问题-方法”知识元的科技文献细粒度抽取与组织方法从底层的知识描述入手，以“背景-问题-方法-目的”为内容主线抽取和组织知识元。该方法深度挖掘了科技文献资源的内容语义和结构价值，为科技文献的细粒度抽取和组织提供新思路，对于缓解科研人员精准的知识获取需求与当前粗放的科技文献服务之间的矛盾具有重要意义。

As the carrier of scientific researchers to record the process of scientific truth verification, experimental observations, research conclusions and other knowledge intelligence, scientific and technical literature is the wisdom crystallization of human scientific and technological development. In the Internet era of information explosion, the accumulation of information resources is not equal to the accumulation of knowledge, and the spurt of scientific and technical literature makes knowledge a more scarce resource. As the essence of scientific research, research users spend a lot of time and energy to locate and integrate the core knowledge points of scientific and technical literature, and the coarse-grained knowledge service of "broad but not specialized" can no longer meet the users' needs for precision and speed. Therefore, how to deeply penetrate into the core contents of scientific and technical literature, integrate the "fragmentation" and systematization of knowledge, and organize the knowledge from surface to point and from point to surface in a three-dimensional way becomes an urgent problem to be solved. In view of this, this paper proposes the fine-grained "problem-method" knowledge element model of science and technology literature with the guidance of knowledge element theory, and on this basis, studies the extraction and organization methods of "problem-method" knowledge elements of science and technology literature, in order to alleviate the gap between users' fine-grained knowledge needs and coarse knowledge services. On this basis, the method of extracting and organizing the "problem-method" knowledge elements of scientific and technical literature is studied, which provides a methodological basis for alleviating the contradiction between users' fine-grained knowledge needs and coarse knowledge services.

The work of this study includes the construction of "problem-method" knowledge elements, knowledge element extraction, multidimensional knowledge organization and experimental validation. (1) Construction of fine-grained "problem-method" knowledge meta-model of scientific and technical literature. The core content elements and auxiliary content elements are extracted from the research questions and research methods of scientific and technical literature, and the external performance and internal characteristics of scientific and technical literature are described in three dimensions, namely, literature resources, content elements and semantic associations, respectively, to construct the semantic description model of "question-method" knowledge elements and realize the systematic description of the research contents of scientific and technical literature. (2) Propose a systemic description of the research content of scientific and technical literature. (2) A knowledge extraction method for scientific and technical literature is proposed for the "problem-method" knowledge meta-model. Based on the syntactic structure and syntactic components of content statements, the knowledge meta-extraction method is designed in two directions: the former introduces markers in the characteristic sentences to enhance the positioning ability of description rules in content element extraction; the latter constructs the content event chain of content statements based on the dependency syntactic analysis and extracts "problem-method" pairs based on semantic continuity. The latter constructs content event chains of content statements based on dependent syntactic analysis, and extracts "problem-method" pairs based on semantic continuity to achieve precise extraction of knowledge elements and relations. (3) A fine-grained organization method of scientific and technical documents based on the "problem-method" knowledge meta-model is proposed. According to the three basic features of vocabulary - semantic, syntactic and semantic, the semantic layer compares the auxiliary content element pairs from two perspectives of common and unique, the syntactic association calculates the internal syntactic structure of the content element pairs by constructing weight assignment rules, the semantic association focuses on the analysis of the conceptual contribution of the content elements themselves and the semantic expansion of the multi-level associated content element sets. The semantic association focuses on the analysis of conceptual contribution of content elements themselves and the semantic range expansion of multi-level associated content elements, and organizes fine-grained knowledge elements from multiple perspectives. (4) Experimentation on knowledge organization of scientific and technical documents based on the "problem-method" knowledge meta-model. The results demonstrate the accuracy of the proposed knowledge element extraction method and the multi-dimensionality and systematicity of the organization method.

The method of fine granularity extraction and organization of scientific and technological documents based on "problem method" knowledge elements starts with the underlying knowledge description, and takes "background problem method purpose" as the main content line to extract and organize knowledge elements. This method deeply explores the content semantics and structural value of scientific and technological literature resources, providing new ideas for fine-grained extraction and organization of scientific and technological literature, and is of great significance in alleviating the contradiction between the precise knowledge acquisition needs of scientific researchers and the current extensive scientific and technological literature services.

[1] MARTINI A, ANNOSI M C, CASPRINI E, et al. Post-acquisition knowledge management practices for exploration and exploitation: insights from a food service organization[J]. Journal of Knowledge Management, 2021, 25(11):129-148.
[2] 刘宁静, 刘音, 王莫言, 等. 数字人文视角下学术名人知识模型构建研究——以李政道数字资源中心为例[J]. 图书情报工作, 2019, 63(23):113-121.
[3] 叶鹰. 质性知识构造与量化知识分析综论:兼论学术对标法理论基础[J]. 中国图书馆学报, 2022, 48(01):38-51.
[4] 王忠义, 夏立新, 李玉海. 基于知识内容的数字图书馆跨学科多粒度知识表示模型构建[J]. 中国图书馆学报, 2019, 45(06):50-64.
[5] 秦春秀, 杨智娟, 赵捧未, 等. 面向科技文献知识表示的知识元本体模型[J]. 图书情报工作, 2018, 62(03):94-103.
[6] 戎军涛. 学术文献内容知识元语义描述模型研究[J]. 情报科学, 2019, 37(07):30-35.
[7] 傅柱, 王曰芬, 徐绪堪, 等. 基于知识元的中文专利文献知识描述框架[J]. 情报理论与实践, 2019, 42(04):145-150.
[8] 王山. 基于OAVU四元组的数值型知识元表示方法与实践[J]. 现代情报, 2022, 42(08):20-27.
[9] 温有奎, 温浩, 徐端颐, 等. 基于知识元的文本知识标引[J]. 情报学报, 2006, 25(3):7.
[10] 李贺, 杜杏叶. 基于知识元的学术论文内容创新性智能化评价研究[J]. 图书情报工作, 2020, 64(01):93-104.
[11] 索传军, 赖海媚. 学术论文问题知识元的类型与描述规则[J]. 中国图书馆学报, 2021, 47(02):95-109.
[12] 秦春秀, 刘杰, 刘怀亮, 等. 基于知识元的科技文本内容描述框架研究[J]. 图书情报工作, 2017, 61(10):116-124.
[13] 化柏林. 学术论文中方法知识元的类型与描述规则研究[J]. 中国图书馆学报, 2016, 42(01):30-40.
[14] 王忠义, 沈雪莹, 黄京. 科技文献资源中方法知识元的抽取研究[J]. 情报科学, 2021, 39(01):13-20.
[15] 徐红霞, 李春旺. 科技文献内容知识点抽取研究综述[J]. 数据分析与知识发现, 2019, 3(03):14-24.
[16] 柴庆凤, 史霖炎, 梅珊, 等. 基于人工特征和机器特征融合的科技文献知识元抽取[J]. 数据分析与知识发现, 2021, 5(08):132-143.
[17] 化柏林. 国内外知识抽取研究进展综述[J]. 情报杂志, 2008, 27(02):3.
[18] 沈雪莹, 欧石燕. 科学文献知识单元抽取及应用研究：梳理与展望[J]. 情报理论与实践, 2022, 45(12):195-207.
[19] 卢超, 章成志, 王玉琢, 等. 语义特征分析的深化——学术文献的全文计量分析研究综述[J]. 中国图书馆学报, 2021, 47(02):110-131.
[20] 安欣, 徐硕, 叶书路, 等. 面向全文本的微观实体抽取及扩散研究[J]. 图书馆论坛, 2021, 41(03):42-49.
[21] WANG S, MAO J, LU K, et al. Understanding interdisciplinary knowledge integration through citance analysis: A case study on eHealth[J]. Journal of Informetrics, 2021, 15(04):101214.1-101214.22.
[22] GABOR K, ZARGAYOUNA H, BUSCALDI D, et al.Semantic Annotation of the ACL Anthology Corpus for the Automatic Analysis of Scientific Literature[C]// Proceedings of the Tenth International Confenrence on Language Resources and Evalution, 2016:3694-3701.
[23] 王忠义, 沈雪莹, 黄京. 基于知识元的中文文本层级分割[J]. 图书情报工作, 2019, 63(07):105-115.
[24] 叶光辉, 彭泽, 陈国梁, 等. 学术文献中的知识单元抽取及其分布特征识别研究[J/OL]. http://kns.cnki.net/kcms/detail/11.1762.G3.20221116.1700.008.html. [2023-03-18].
[25] 石义金, 王忠义, 沈雪莹, 等. 基于序列模式的科技文献中知识元抽取研究[J]. 情报理论与实践, 2020, 43(11):144-149.
[26] 徐康, 余胜男, 陈蕾, 等. 基于语言学知识增强的自监督式图卷积网络的事件关系抽取方法[J/OL]. http://kns.cnki.net/kcms/detail/10.1478.g2.20220920.1551.002.html. [2023-03-18].
[27] LIAKATA M, SAHA S, DOBNIK S, et al. Automatic recognition of conceptualization zones in scientific articles and two life science applications[J]. Bioinformatics, 2012, 28(7):991-1000.
[28] DUCK G, KOVACEVIC A, ROBERTSON D L, et al. Ambiguity and variability of database and software names in bioinformatics[J]. Journal of biomedical semantics, 2015, 6(01):29.
[29] 陈鹏, 郭剑毅, 余正涛, 等. 融合领域知识短语树核函数的中文领域实体关系抽取[J]. 南京大学学报(自然科学), 2015, 51(01):181-186.
[30] LI P, YUAN Z, TU W, et al. Medical Knowledge Extraction and Analysis from Electronic Medical Records Using Deep Learning[J]. Chinese Medical Sciences Journal, 2019, 34(2):70.
[31] 曹树金, 曹茹烨. 基于知识图谱支持科研创新的跨学科知识发现研究[J]. 情报理论与实践, 2022, 45(11):10-20.
[32] 曹思源, 马海云. 领域知识组织理论基础及方法分类简述[J]. 情报资料工作, 2021, 42(05):14-22.
[33] 赵志耘, 刘耀, 朱礼军, 等. 复杂信息环境下知识组织和再利用模式与方法研究[J]. 情报学报, 2022, 41(12):1266-1279.
[34] ZHAO H, PAN Y, YANG F. Research on Information Extraction of Technical Documents and Construction of Domain Knowledge Graph[J]. IEEE Access, 2020, 8:168087-168098.
[35] 任亮, 杜薇薇, 刘伟利. 面向科技文献知识元的知识图谱构建研究[J]. 情报科学, 2022, 40(09):26-31.
[36] 刘杰, 秦春秀, 赵捧未, 等. 基于知识元的科技文本资源内容组织方法[J]. 情报理论与实践, 2018, 41(04):128-133.
[37] 庄文杰, 谈国新, 侯西龙等. 非物质文化遗产视频知识元组织模型研究[J]. 情报科学, 2018, 36(12):25-32.
[38] 王忠义, 郑鑫, 王珂莹. 面向用户生成内容的多粒度知识组织研究[J]. 情报学报, 2022, 41(10):1034-1043.
[39] 曾刚, 赵雪芹. 基于知识元的万里茶道数字资源知识抽取与组织研究[J]. 情报理论与实践, 2021, 44(10):173-178+164.
[40] 董坤. 基于知识元的非物质文化遗产知识抽取与组织研究[J]. 情报理论与实践, 2021, 44(09):155-160+148.
[41] 陈果, 彭家彬, 肖璐. 基于“问题—方法”知识抽取的科研领域知识演化研究：以人工智能为例[J]. 情报理论与实践, 2022, 45(06):32-38.
[42] 温有奎, 温浩, 乔晓东. 让知识产生智慧——基于人工智能的文本挖掘与问答技术研究[J]. 情报学报, 2019, 38(07):722-730.
[43] 王忠义, 涂悦, 夏立新. 科技文献资源中学科知识漂移研究[J]. 情报理论与实践, 2021, 44(06):118-124.
[44] 黄丹俞. 设计科学研究范式；架起阅读推广理论与实践的桥梁[J]. 图书与情报, 2016(03):61-66.
[45] 冯儒佳, 王忠义, 王艳凤, 等. 科技论文的多粒度知识组织框架研究[J]. 情报科学, 2016, 34(12):46-50+54.
[46] YAO Y, ZHAO L. A measurement theory view on the granularity of partitions[J]. Information Sciences: An International Journal, 2013, 213:1-13.
[47] 王凯, 潘玮, 杨枢, 等. 基于模糊概念格的丁香园社区用户多粒度画像研究[J]. 情报理论与实践, 2020, 43(08):103-111.
[48] 徐绪堪, 房道伟, 蒋勋, 等. 知识组织中知识粒度化表示和规范化研究[J]. 图书情报知识, 2014(06):101-106+90.
[49] IMRE H. The epsilon-knowledge:an emerging complement of Machlup's types of disciplinary knowledge[J]. Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 2022, 36(18):1-18.
[50] 于秀慧, 李宝山. 基于知识元的知识管理[J]. 山东图书馆学刊, 2013(01):10-13.
[51] 孙成江, 吴正荆. 知识服务战略；创建增值联盟[J]. 情报科学, 2002(10):1028-1029+1035.
[52] 温有奎, 徐国华. 知识元链接理论[J]. 情报学报, 2003, 22(6):6.
[53] 索传军, 戎军涛. 知识元理论研究述评[J]. 图书情报工作, 2021, 65(11):133-142.
[54] 索传军, 盖双双. 知识元的内涵、结构与描述模型研究[J]. 中国图书馆学报, 2018, 44(04):54-72.
[55] 高劲松, 付家炜. 基于知识元的文物图像资源细粒度知识表示方法研究[J]. 情报科学, 2022, 40(01):16-24.
[56] 叶鑫, 董路安, 宋禺. 基于大数据与知识的“互联网+政务服务”云平台的构建与服务策略研究[J]. 情报杂志, 2018, 37(02):154-160+153.
[57] 赵雪芹, 李天娥, 曾刚. 面向数字人文图像资源的知识元本体构建及关联展示研究[J]. 情报理论与实践, 2022, 45(09):180-187.
[58] 武建光, 苏云梅, 于琦, 等. 基于知识元的学术文献分类研究[J]. 情报理论与实践, 2019, 42(03):160-165.
[59] 黄容, 何杨煜琪, 王忠义, 等. 数字图书馆数值知识元检索系统设计[J]. 图书情报工作, 2018, 62(14):125-132.
[60] LI X, LIU H, ZHAO X, et al. Automatic approach for constructing a knowledge graph of knee osteoarthritis in Chinese[J]. Health Information Science and Systems, 2020, 8(01):1-8.
[61] 付哲, 李军. 高性能正则表达式匹配算法综述[J]. 计算机工程与应用, 2018, 54(20):1-13.
[62] KLEENE S C. Representation of Events in Nerve Nets and Finite Automata[R]. Santa Monica, CA:RAND Project Air Force, 1951.
[63] THOMPSON K. Programming Techniques: Regular expression search algorithm[J]. Communications of the ACM, 1968, 11(06):419-422.
[64] 王贵荣, 饶高琦, 荀恩东. 基于大规模语料库的现代汉语动宾搭配知识库构建[J]. 中文信息学报, 2021, 35(01):34-42+53.
[65] 周国光. 汉语配价语法论略[J]. 南京师大学报(社会科学版), 1994(04):103-106+121.
[66] 黄昌宁, 苑春法, 潘诗梅. 语料库、知识获取和句法分析[J]. 中文信息学报, 1992(03):1-6.
[67] 万齐智, 万常选, 胡蓉, 等. 基于句法语义依存分析的中文金融事件抽取[J]. 计算机学报, 2021, 44(03):508-530.
[68] 王步康, 王红玲, 袁晓虹, 等.基于依存句法分析的中文语义角色标注[J]. 中文信息学报, 2010, 24(01):25-29+47.
[69] TAUFIQ U, PULUNGAN R, SUYANTO Y. Named entity recognition and dependency parsing for better concept extraction in summary obfuscation detection[J]. Expert Systems With Applications, 2023, 217.
[70] ZHAI P, HUANG X, ZHANG B, et al. Relation Extraction Based on Fusion Dependency Parsing from Chinese EMRs[J]. Scientific Programming, 2020, 2020(02):1-9.
[71] YIN Z, SHAO J, HUSSAIN M J, et al. DPG-LSTM: An Enhanced LSTM Framework for Sentiment Analysis in Social Media Text Based on Dependency Parsing and GCN[J]. Applied Sciences, 2022, 13(1):354.
[72] SHOU Y, MENG T, AI W, et al. Conversational emotion recognition studies based on graph convolutional neural networks and a dependent syntactic analysis[J]. Neurocomputing, 2022, 501:629-639.
[73] YANG Y, LIN H, YANG Z, et al. ADPG: Biomedical entity recognition based on Automatic Dependency Parsing Graph[J]. Journal of biomedical informatics, 2023, 140:104317-104317.
[74] POSADAS-DURÁN JP, SIDOROV G, GÓMEZ-ADORNO H, et al. Algorithm for Extraction of Subtrees of a Sentence Dependency Parse Tree[J]. Acta Polytechnica Hungarica, 2017, 14(03):79-98.
[75] 李葆嘉, 刘慧. 西方语义学传入中国的路径[J]. 汉语学报, 2018, 61(01):2-11+95.
[76] 周建设. 语义学的研究对象与学科体系[J]. 首都师范大学学报(社会科学版), 2000(02):87-92.
[77] 伍廉光. 语义学导论[M]. 湖南: 湖南教育出版社, 1988.
[78] 莱昂斯, 汪榕培. 语义学引论[M]. 北京: 外语教学与研究出版社, 2000.
[79] 牟冬梅. 数字图书馆知识组织语义互联策略及其应用研究[D]. 吉林: 吉林大学, 2009.
[80] 祁天娇, 冯惠玲. 档案数据化过程中语义组织的内涵、特点与原理解析[J]. 图书情报工作, 2021, 65(09):3-15.
[81] MUPPAVARAPU V, RAMESH G, GYRARD A, et al. Knowledge extraction using semantic similarity of concepts from Web of Things knowledge bases[J]. Data & knowledge engineering, 2021(Sep.):135.
[82] LI H, HONG T. A semantic ontology for representing and quantifying energy flexibility of buildings[J]. Advances in Applied Energy, 2022, 8.
[83] FORMICA A, TAGLINO F. Semantic relatedness in DBpedia: A comparative and experimental assessment[J]. Information Sciences, 2023, 621:474-505.
[84] ROCHA E S S, FERREIRA C F, ARAKAKI A C S. Metadata and Semantics Metadata: Contemporary Discussions in the Digital Context[J]. Journal of Library Metadata, 2022, 22(1-2):149-166.
[85] AZAD HK, DEEPAK A, AZAD A. LOD search engine: A semantic search over linked data[J]. Journal of Intelligent Information Systems, 2021, 59(01):71-91.
[86] 陶俊. 词表语义组织研究的演进(1998-2018)[J]. 图书情报工作, 2018, 62(21):140-148.
[87] 石湘, 刘萍. 基于知识元语义描述模型的领域知识抽取与表示研究——以信息检索领域为例[J]. 数据分析与知识发现, 2021, 5(04):123-133.
[88] 林定夷. 科学问题与科学目标[J]. 中国社会科学, 1991(05):29-38.
[89] 李跃艳, 王昊, 邓三鸿, 等. 面向事件本体的医学文本语义关联化研究[J]. 情报学报, 2022, 41(05):497-511.
[90] 王末, 崔运鹏, 陈丽, 等. 基于深度学习的学术论文语步结构分类方法研究[J]. 数据分析与知识发现, 2020, 4(06):60-68.
[91] 丛靖怡, 艾文华, 胡广伟. 基于信息交互视角的突发公共事件舆情反转治理路径研究[J]. 情报学报, 2022, 41(06):594-608.
[92] NEMETH, T E, BIBOK K. Interaction between grammar and pragmatics: The case of implicit arguments, implicit predicates and co-composition in Hungarian[J]. Journal of Pragmatics, 2010, 42(02):501-524.
[93] 何自然, 冉永平. 关联理论—认知语用学基础[J]. 现代外语, 1998(03):95-109+94.
[94] JACCARD P. The distribution of the flora in the alpine zone[J]. New Phytologist, 1912, 11(02) :37-50.
[95] 陆丙甫. 从语义、语用看语法形式的实质[J]. 中国语文, 1998(05):353-367.
[96] 秦春秀, 郑梦悦, 马续补, 等. 基于智能主题图的科技文献细粒度知识组织模型[J]. 情报杂志, 2021, 40(11):169-175+135.
[97] 冯志伟. 汉语词组型术语的结构[J]. 科技术语研究, 2004(02):35-37.
[98] 杨泉. 一种基于决策树和词义相似度的N1+N2结构语法关系判定方法[J]. 计算技术与自动化, 2021, 40(04):36-41.
[99] 苗夺谦, 王珏. 粗糙集理论中概念与运算的信息表示[J]. 软件学报, 1999(02):2-5.
[100] 李思旭. 从语言类型学看三个平面互动研究[J]. 汉语学习, 2014, 200(02):20-31.
[101] YEPES A J, MORK J, ARONSON A. Using the Argumentative Structure of Scientific Literature to Improve Information Access[C]// Association for Computational Linguistics. Proceedings of the 2013 Workshop on Biomedical Natural Language Processing, August 8, 2013. Sofia, Bulgaria. BioNLP 2013, 2013:102-110.