网络特征对文献间接关联发现的评价作用

摘要

New scientific hypotheses detected by mining the potential indirect association inliterature according to the studies on literature-based knowledge discovery are increasinglyapplied in biomedical field and evaluation of literature-based indirect association discovery is a hot spot in recent studies on literature-based knowledge discovery . The role of network characteristics in evaluation of literature-based indirect association discovery during the litera-ture-based knowledge discovery was thus studied.The new indexes for evaluaing the literature-based indirect asso-ciation discovery were esatablished by integrating the co-ocurrent statistic information and the network charateris-tics, which are of greatimportance for improving literature-based knowledge discovery and constructing knowledge discovery system .%基于文献的知识发现研究通过文献中潜在的间接关联的挖掘获取新的科学假设，在生物医学领域已经得到了越来越广泛的应用。间接关联性的评价是当前文献知识发现研究的一个热点问题；本文主要探讨网络特征在文献知识发现过程中间接关联的评价作用，整合共现统计信息与网络结构特征，建立了间接关联评价的新指标，这对于提高文献知识发现效率与构建知识发现系统具有重要意义。