学术文本词汇功能识别--在论文新颖性度量上的应用  

Application of Lexical Functions in Novelty Measurement of Academic Papers

作　　者：罗卓然 陆伟[1,2] 蔡乐 程齐凯[1,2] Luo Zhuoran;Lu Wei;Cai Le;Cheng Qikai(School of Information Management,Wuhan University,Wuhan 430072;Information Retrieval and Knowledge Mining Laboratory,Wuhan University,Wuhan 430072)

机构地区：[1]武汉大学信息管理学院,武汉430072 [2]武汉大学信息检索与知识挖掘研究所,武汉430072

出　　处：《情报学报》2022年第7期720-732,共13页Journal of the China Society for Scientific and Technical Information

基　　金：国家社会科学基金重大项目“基于认知计算的学术论文评价理论与方法研究”(17ZDA292)。

摘　　要：为进一步挖掘学术论文新颖性的丰富内涵,本文以组合创新理论为基础,开展了基于词汇功能的学术论文新颖性度量研究。以ACM(Association for Computing Machinery)Digital Library收录的论文为数据,提出了面向CS(computer science)领域进一步预训练的词汇新颖性计算方法和基于语义相似度的问题-方法组合新颖度计算流程,分别计算了问题词、方法词、问题-方法组合和论文的语义新颖性,并将本文语义新颖性计算方法与已有的词频共现新颖性计算方法进行了对比。研究结果表明,ACM Digital Library收集的论文在研究方法和研究问题上创新度均较高,相较于已有的论文新颖性计算方法,本文提出的方法能从语义层面捕获更为精细的新颖性差异。To further investigate the novelty measure of academic papers,we conducted a study on the novelty measurement of academic papers based on lexical functions.Specifically,we propose a method to calculate the novelty of words for further pre-training in the field of computer science based on the data in papers in Association for Computing Machinery(ACM)Digital Library and present a novelty calculation process for“question-method”combinations.We calculated the novelty of the test set data and performed a comparison between our method and existing methods.Results show that the papers collected from the ACM database are more innovative in terms of both research methods and research questions,and the method proposed in this paper can capture more fine-grained novelty differences at the semantic level.

关 键 词：新颖性度量 词汇功能 问题-方法组合 预训练模型 

分 类 号：G254[文化科学—图书馆学]