基于分析工具与期刊同现双重优化的生物医学领域学科分类体系映射研究

Semantic Mapping Between Biomedical Discipline Classifications Based on Analysis Tools and Journal Co-occurrence

目的:针对国内外主要学科分类体系不统一、缺少语义关联等问题,以Web of science、Scopus分类体系与教育部《学科目录》映射为例,通过分析不同分类体系结构及期刊分类数据,运用多种语义映射与优化方法建立相关类目之间的关联关系,以期实现分布式科技文献资源的统一组织,提升科研成果一站式检索、分类导航与统计评价效果。方法:采用基于中心结构模型的语义映射方法,初步构建Web of science、Scopus生物医学相关类目与教育部《学科目录》之间的关联映射表;采集Web of science、Scopus共同收录的生物医学期刊分类数据,基于Incites学科分析工具与期刊同现两种方法完成映射结果的优化与验证。结果:分别建立了Web of science 103个、Scopus 153个类目与教育部《学科目录》中12个生物医学类目之间126对、200对语义映射关系。结论:研究成果为我国部分高校和科研院所开展生物医学相关学科评估奠定基础,并为其他学科领域的分类体系映射提供理论与实践参考。

Objective:Due to the lack of consistency and interoperability of dominant discipline classifications,the semantic mapping between Web of science(WOS) category,Scopus and Chinese discipline catalogue was taken as an example for demonstration research,based on several mapping methods and optimization with the aids of classified journal data.Methods:The hub model was adopted for basic semantic mapping,followed by the double optimization processing dependent on annotated data from Incites analysis tools and co-occurred biomedical journals.Results:126 and 200 pairs of semantic mapping were established between 103 of WOS category,153 of Scopus and 12 biomedical items of Chinese catalogue separately.Conclusion:The mapping results will support the discipline evaluation of biomedical colleges and universities in China,as well as provide reference to semantic mapping in other fields.