Abstract:

Biological breeding science has experienced full speed development for over half a century, and the number of scientific and technical literature has increased rapidly. Through analyzing the knowledge development in subject areas, this paper provided the most up\|to\|date progress of varied discipines, detected research hot spots, and discovered the internal connection between knowledge and their integration, so as to forecast the tendency for future development of various disciplines. A knowledge evolution model was constructed based on various phenomena in knowledge evolution process. This model can automatically analyze the relationships, properties, change degree and changing trends between different time windows. Biological breeding documents published between 1995-2012 were selected, combined with co\|author in the network and country cooperation networks to analyze the knowledge evolution phenomena, displayed all results under different time windows, and preliminary confirmed the feasibility of time slice strategy on seeking knowledge evolution phenomenon.

Key words: knowledge evolution, topic evolution, co\, word analysis, biological breeding, cooperation network

摘要：

生物育种科学经过半个多世纪的飞速发展，科技文献数量迅猛增加，通过对学科领域知识的发展分析，详细了解领域内各学科的发展状态，探测研究热点，发现知识间的扩散和融合，从而为学科领域未来的发展趋势进行更好地预测。依据知识演化过程中产生的各种现象设计了知识演化方法模型，该模型可以对不同时间窗聚类间的主题关系、关系性质、变化程度及其所代表的演变趋势进行自动分析。选取1995-2012年生物育种领域的文献，结合共著网络和国家合作网络，重点对知识演化现象进行了分析，展示了不同时间窗聚类结果，初步证实了时间窗划分用于探寻知识演化现象的可行性。

关键词: 知识演化；主题演化；共词分析；生物育种；合作网络