中文摘要：

随着专利侵权案件的频繁发生，识别被侵权可能性较高的专利对企业规避专利侵权风险有着重要意义。以LED照明专利为实例，用指标显著性检验和因子分析法筛选LED照明领域诉讼专利特征，建立高诉讼风险专利的逻辑回归识别模型。对LED照明高诉讼风险专利在分支技术领域的分布情况和重要企业间的分布情况做统计分析，识别出可能发生诉讼纠纷的分支技术和可能主动发起诉讼的企业，辅助企业制定有效的专利诉讼策略。

英文摘要：

With the frequent occurrence of patent infringement, recognizing potential litigation patent is of great significance for enterprises to avoid the risk of patent infringement. In this paper, we take LED lighting as the case. Significance test and factor analysis have been a- dopted to identify the characteristics of litigation patent. We establishe a potential litigation patent model based on Logistic regression. Then statistical analyses are applied to present the layout of technical area and key enterprises. Then techniques which are easily involved in patent litigation and enterprises, which intend to file a lawsuit are detected, assisted in developing litigation strategies for enterprises.