基于突发词检测的科研热点发掘服务模型研究  

On Scientific Research Hotspot Mining Service Model in Libraries

作　　者：王征[1] 易莉[1] 赵磊[1] 

机构地区：[1]西南财经大学经济信息工程学院&研究生院,成都611130

出　　处：《情报杂志》2015年第12期176-180,共5页Journal of Intelligence

基　　金：国家社会科学基金青年项目"群体性事件管理推演与应对措施验证研究"(编号:14CGL050)

摘　　要：科研热点发掘及预测是图书馆信息服务领域的研究重点之一。为应对科技期刊服务工作中热点挖掘中的数据量大、信息来源单一、时滞较大等问题,提出了基于突发词检测的科研热点发掘服务模型。该模型基于灰色关联理论构建,从海量的关键词中检测和识别突发词,并对其进行跟踪和排序,最终发现其所在领域的研究前沿和变迁轨迹。仿真实验证明,较之图谱式发现模型,该模型具有较高的科研热点发掘精确度、关键词覆盖度,适用于图书馆科研热点服务的热点发掘及预测工作。The scientific research hotspot exploration and prediction is a key research topic of library information service works. In order to deal with the problems of mass data, single information source and lengthy time lag in hotspot mining works, a scientific research hots- pot mining service model was presented based on hursty topic detection. And the model was constructed based on grey relational theory. It could detect and recognize the bursty topics from mass keywords, thus facilitating tracking and sorting the bursty topics on the scientific re- search front. Simulation results show that the new structure and methods make it possible for the service model to provide comprehensive, personalized and accurate scientific research hotspot mining service for editors. And it can provide practical tools and references for the sci- entific and technical periodical library services.

关 键 词：信息服务 科研热点 科技期刊 灰色关联 突发词 

分 类 号：TP301[自动化与计算机技术—计算机系统结构] G253[自动化与计算机技术—计算机科学与技术]