摘要
[目的/意义]针对海量科技文献新颖性评估需求,从科技论文研究问题与方法视角出发,利用科技论文摘要与相关引文文本,构建科技论文的评述画像,用于辅助科技论文新颖性评估。[方法/过程]提出一种综合“自评+他评”内容的科技论文评述画像方法。该方法基于期刊论文结构化摘要文本解析问题和方法等句子,以单篇期刊论文为目标文献,将施引文献中引文句子作为引文数据集并分类后,可视化呈现科技论文的自评和他评内容。目前科技论文评述画像分为两种:一是按照引用问题与方法等分类并可视化,二是按照自引他引及情感分析进行分类可视化。[结论/结果]构建的科技论文评述画像能够从论文作者与引文作者的双重视角综合展示目标论文的研究成果,结合自引他引及情感分析信息可有效辅助科技论文新颖性评估。
[Purpose/significance] To assist the novelty evaluation of massive scientific literature, article abstracts and relevant citation texts are used to construct the commentary profile of scientific article from the perspective of research question and research method.[Method/process] This paper proposes a method of review profiles of scientific article that integrates the content of “self-evaluation + other-evaluation”.This method is based on sentences such as analytical questions and methods of structured abstracts of journal article, takes a single journal article as the target article, takes citation sentences in citing articles as a citation data set and categorizes them according to rules, and visualizes the self-evaluation and other-evaluation.There are two kinds of profiles in the construction of review portraits in scientific papers: one is visual classification according to problems and methods in citation texts, and the other is visual classification according to self-citation, other-citation and sentimental analysis.[Result/conclusion] The review profile of scientific article can display research results of the target article combining the perspectives of article authors and citation authors, which could assist the novelty assessment of scientific articles by combining the information of self-citation, other-citation and sentimental analysis.
出处
《情报理论与实践》
北大核心
2023年第1期159-167,共9页
Information Studies:Theory & Application
基金
国家重点研发计划项目“颠覆性技术识别理论、方法与专家预判系统”(项目编号:2019YFA0707201)
中国科学技术信息研究所创新研究基金项目“科技论文原创性与新颖性评估方法研究”(项目编号:MS2022-05)的成果。
关键词
科技论文
新颖性评估
引用分类
评述画像
可视化
scientific article
novelty evaluation
citation classification
review profile
visualization