基于引文关联数据的引文路径识别和引文网络可视化研究

Citation Path Recognition and Citation Network Visualization Based on Citation Linked Data

【目的/意义】对CSSCI"引文分析"领域的引文题录数据进行关联数据发布实验,利用可视化软件RelFinder和Graphviz进行多角度的引文路径识别和可视化研究,以期利用关联数据技术来揭示不同学术文献之间的引用关系和潜在规律。【方法/过程】首先,在CSSCI中以"引文分析"为检索词进行检索,得到557篇被引文献构成引文数据集,同时抽取所对应的1105篇施引文献构成施引文献数据集,共生成1598条引用关系。接着,构建轻量级任务型引文本体对文献的题录数据和引用关系进行规范化描述,并利用开源软件OpenLink Virtuoso将引文数据发布为引文关联数据。最后,使用RDF可视化软件RelFinder进行学术文献之间的引用关系发现和关键引文路径识别研究,利用Graphviz进行引文网络的绘制和可视化呈现。【结果/结论】关联数据技术能较好地描述和揭示学术文献之间的引用关系,RelFinder和Graphviz在引文路径识别和引文网络可视化方面具有较高的实践意义。

【Purpose/significance】In this paper, part of the citation bibliographic data in the Chinese Social Sciences Citation Index was converted into linked data. Then we carried out some citation path recognition and visualization experiments from different aspects with RelFinder and Graphviz, hoping to reveal the citation relations and some potentialcitation laws between different academic units.【Method/process】First of all, We used 'citation analysis' as a search termto retrieve the CSSCI, and obtained 557 cited papers and their corresponding 1105 citing papers in total. These papers wereused to compose the cited data set and the citing data set. We also extracted 1598 citation relations between citing papersand cited papers. Then, a lightweight task-based citation ontology was constructed to normalize the CSSCI citationbibliographic data. After that, we converted the citation data into RDF triples and used OpenLink Virtuoso to publish RDFtriples as citation linked data. Finally, the RDF visualization software RelFinder was used to find the citation relationbetween different academic units, and Graphviz was used to draw the citation networks of papers, journals and authors.【Result/conclusion】The technology of linked data could describe and reveal the citation relationship between academicliterature. RelFinder and Graphviz have an important practical significance in citation path recognition and citation network visualization.