中文摘要：

通过对从美国专利商标局（ The US Patent and Trademark Office,USPTO）网站中下载的某领域专利进行整理,得到初始统计数据。将USPTO专利子类作为技术节点,技术之间共现关系作为边,建立技术的共现网络,再利用复杂网络中的“k-核”分析方法对共现网络进行分析,进行技术筛选。提出技术识别的层次性评估指标,运用证据推理对筛选出的技术进行评估,选出核心技术。最后以激光切割技术为例,展示和验证了此方法。

英文摘要：

Based on an analysis of the original data of certain field collected from The US Patent and Trademark Office （ USPTO） web-site, the patent subclasses were treated as nodes and the co-occurrence relations were considered as edges in the constructed technology co-occurrence network. By applying the complex network method of ＂k-core＂, the co-occurrence network was analyzed and techniques were refined. The hierarchical evaluating indexes proposed were then applied to evaluating the techniques refined, and core techniques were recognized at last. The method proposed in this paper was verified by an example of laser cutting techniques.