中文摘要：

共词网络分析是科学计量学研究领域中的重要方法，同时也是文献内容分析的常用方法。本文主要从关键词共词网络演化角度进行了学科知识结构的可视化展现、发展趋势的预测以及研究前沿和热点辨识研究。介绍了作者关键词共现网络和机标关键词共现网络两种典型学科共词网络并加以比较，得出机标关键词共现网络更能清晰的展现学科知识结构及演化历程的结论。在熵理论共现网络的实证研究中，本文收集了自然科学和社会科学1990～2014年WoS平台中有关熵理论的科学文献，并分别做了熵科学文献的作者关键词共词网络演化图和熵机标词共词网络演化图，对社会科学类科学文献分别做了作者关键词网络演化图和机标关键网络演化图谱，比较分析了熵理论在各种共词网络图中的泛化与创新路径，以此发现了熵理论的研究前沿、热点以及熵理论的演化机理，这对未来科学发展的创新有一定借鉴作用。

英文摘要：

The co - words network analysis is an important method in the field of scientometrics, and also the commonly used method in the literature content analysis. In this paper, this method is used displaying the discipline knowledge structure, forecasting the development trend and identifying the frontier and hot spot from the perspective of co - word networks evolution. The authors＇ co -words network and the keyword plus networks were introduced and compared, and the result show that the latter display subject knowledge structure and evolution path more clearly. In empirical study of the entropy theory co- word networks, the scientific literatures in natural science and social science from 1900 to 2014 were collected from the WoS platform, the authors＇ eo - word network and the keyword plus network were built, in the social science the entropy author keywords evolution maps and the keywords plus evolution maps were respectively made. In these maps, after the innovation path of entropy theory analyzed and compared, the research front, hot loci and the evolution mechanism of the entropy theory were found, which has an important guiding about the scientific innovation and development in the future.