基于SAO结构语义分析的新兴技术识别研究

Identifying Emerging Technologies Based on Subject – Action-Object

[目的/意义]如何在新技术不断涌现的技术大海中寻找和识别有颠覆性潜力的新兴技术,对于企业或国家来说是至关重要的战略议题。针对目前在新兴技术识别中存在的不足,即专利引用信息的滞后性和关键词聚类既不能深入反映专利文献表达的主题,也不能很好地揭示技术主题之间的关联关系等问题。[方法/过程]研究提出了基于SAO结构语义分析法的新兴技术识别模型。并首先利用文本挖掘方法抽取专利权利要求项中的Subject–Action-Object(SAO)结构,其次利用改进的语义相似度算法对专利文本进行聚类,然后结合基于时间切片的专利地图和语义分析法识别新兴技术。[结果/结论]用钙钛矿太阳能电池作为实证进行了研究,验证了该模型的可行性和有效性。

[Purpose / Significance] Howto detect and identify the emerging technologies with commercial potentials in the newtechnologies sea is a crucial issue for the national RD strategic planning of China and enterprises practices. [Method / Process] As for avoiding the weaknesses of using patent citation and keywords analysis approach to identify the emerging technologies,this paper proposes a method to identify emerging technologies based on Subject-Action-Object（ SAO） structures. Firstly,the paper uses the data mining method to extract the SAO structures from the claims of patents. Then,the patent clustering topics are obtained by applying the semantics similarity analysis method. After that,emerging technologies are identified by using patents map and SAO structures analysis method. [Result / Conclusion] Finally,the perovskite solar cells technology domain case is used to analyze the effectiveness and feasibility of the method.