摘要
从海量科技文献中自动挖掘隐含主题、研究人员的研究兴趣及其演化规律是信息服务迈向知识服务需要解决的关键问题之一。目前的方法多从静态的角度分析文献主题、科研人员的研究兴趣,而演化分析的方法主要集中文档的内部特征,即文档内容本身,很少考虑作者等外部特征。基于此,本文在AT和ToT模型的基础上构建了作者主题演化(AToT)模型,并给出了一种估计AToT模型参数的吉布斯采样方法。该模型集成了AT和ToT模型的优势,不仅可以揭示科技文献中隐含的主题、作者的研究兴趣,而且可以挖掘研究兴趣随时间变化的规律。最后,以1740篇NIPS会议论文集作为实验数据,通过与AT模型的对比分析验证了AToT模型的可行性和有效性。
One of the key problems in upgrading information services towards knowledge services is to automatically mine latent topics, researchers' interests and their evolution patterns from large-scale scientific & technical literatures. Most of current methods are devoted to discover static latent topics and research interests. Nevertheless, previous evo]ution analysis research mainly focuses on analyzing intra-features of documents, namely documents' text content without considering directly extra-features of documents such as authors. To overcome this problem, on the basis of Author-Topic (AT) model and Topics over Time (TOT) model, Author-Topic over Time (AToT) model is constructed in this study, and Gibbs sampling method is utilized to estimate corresponding parameters. This model is not only able to discover latent topics and researchers' interests, but also to mine their changing patterns over time. Another way to say this is that our AToT model combines the advantages of AT and ToT models. Finally, the extensive experimental results on NIPS dataset with 1740 papers indicate that our AToT model is feasible and efficient.
出处
《情报学报》
CSSCI
北大核心
2013年第9期912-919,共8页
Journal of the China Society for Scientific and Technical Information
基金
“十二五”国家科技支撑计划“面向外文科技知识组织体系的大规模语义计算关键技术研究”(2011BAH10B04)
“基于STKOS的知识服务应用示范”(2011BAH10B06)
中国科学技术信息研究所预研项目“基于词系统的领域深层主题规律揭示分析研究”(YY201216)资助