摘要
研究亮点是具有宣传功能的特殊文本,概括了一篇学术论文研究方法和结论中的要点,亮点的认知和应用有利于推进知识交流和科研创新。文章以ScienceDirect数据库564篇论文亮点作为数据样本,探索亮点的内外部特征,采用多维度分析法对亮点进行语言特征分析,阐述亮点和摘要的语言特征差异,分析论文被引次数与亮点语言特征的关系,基于此提出亮点写作建议。同时,结合人工标注,借助LDA模型和VOSviewer工具分别从整体和分类两个维度识别亮点文本主题。研究结果发现:亮点整体上具有较强的信息性和非叙述性,指称表达较清晰,交互性和显性劝说性较弱,客观程度和精确度相对较高;被引次数高的论文亮点更倾向于使用较多的基数词、数量词、强调语和独立并列从句;作者撰写亮点时,可以更多地展示具体数据和数量关系;亮点可以分为方法型亮点、结论型亮点以及少量的其他型亮点,其内容能够反映所在科研领域的关键信息、创新性方法前沿和主题重点。
Research highlights,special texts with a promotional function that outline the research methods and conclusions of an academic paper,can facilitate the advancement of knowledge exchange and scientific innovation.Based on 564 paper highlights from the ScienceDirect database,this paper explores the internal and external characteristics of research highlights,analyzes its linguistic features by means of multidimensional analysis(MDA) method,elaborates on the differences between the linguistic features of highlights and abstracts,examines the relationship between the citations of papers and the linguistic features of highlights,and proposes some suggestions for highlight-writing.With the help of manual annotations,the paper identifies the topics of the highlights from an integrated and a categorized point of view respectively,using the LDA model and the VOSviewer tool.The results show that:highlights are generally more informative and non-narrative,with clearer expressions,less interactive and explicit persuasiveness,and a relatively high degree of objectivity and precision;highlights of highly cited papers tend to use more cardinal numbers,quantifiers,emphases and independent or coordinate clauses,demonstrating specific data and their correlation;highlights can be divided into method-based highlights,conclusion-based highlights and(to a lesser extent)other highlights,which indicate the key information of a scientific field,unfolding the innovative methods and focuses in the field.
作者
杨思洛
程濛
莫莹莹
YANG Siluo;CHENG Meng;MO Yingying
出处
《图书馆论坛》
CSSCI
北大核心
2023年第7期26-37,共12页
Library Tribune
基金
国家社会科学基金后期资助项目“基于全文计量分析的知识交流体系研究”(项目编号:22FTQB003)研究成果。
关键词
学术论文
研究亮点
语言特征
多维度分析
主题识别
research article
research highlights
linguistic features
multidimensional analysis
topic identification
作者简介
通信作者:杨思洛,58605025@qq.com,武汉大学信息管理学院教授、博士生导师;程濛,莫莹莹,武汉大学信息管理学院图书馆学系硕士研究生。
