中文摘要：

按照德温特分类号所代表的301个行业,将随机选取的28家500强样本企业发表的专利进行归类,利用数据挖掘、社会网络分析、信息可视化等技术手段,计算专利的相关系数和绘制专利相关性网络,分析样本企业的技术集群和技术相似性,研究关键企业在产业之间技术联系的中介作用。

英文摘要：

According to 301 industries represented by Derwent Class Codes,patents granted to 28 enterprises from Global 500 are selected randomly and classified.By using data mining,social network analysis（SNA）,information visualization,we map the patent correlation network after calculating the correlation coefficients.The procedure is useful in identifying technical clusters and analyzing their technology proximity,as well as the mediating roles of the key enterprises among different industries.