[目的/意義]為了在大規模的專利文獻中,面向技術需求實現專利文獻和技術需求的匹配,識別潛在的技術轉移專利,以提高技術轉移信息服務水平。[方法/過程]文章基于語義TRIZ標注專利文獻和技術需求的技術問題、技術功能、技術效果三種技術特征詞組,通過同義詞典計算特征間的相似性,構建語義的文本—詞項向量空間模型,計算專利文獻和技術需求的相似性,根據相似度排序識別潛在的技術轉移專利。并以新能源汽車電池領域為研究對象進行實證研究。[結果/結論]實證結果表明,針對技術需求,在一定程度上能識別與其匹配的專利文獻,識別潛在的技術轉移專利。[局限]本研究的局限性在于抽取的技術特征詞組數量有限,未來將結合關鍵詞和技術特征詞組更加全面地表示專利文獻和技術需求。

[Purpose/significance] In order to match the patents and the technology demand oriented technology demand in the large-scale patents,identify potential technology transfer patents,and improve the level of technology transfer information services.[Method/process] This study labels three technical feature phrases of problem,function and effect for patents and technology demands based semantic TRIZ.Through calculating the similarity between features by the synonym dictionary,constructing a semantic text-items vector space model,this study calculates the similarity between patents and technology demand,and identifies potential technology transfer patents according to similarity ranking.This study takes the field of new energy vehicle battery as the research object for empirical analysis.[Result/conclusion]The result shows that,in view of the technology demand,it can identify the matching patents to a certain extent and identify potential technology transfer patents.[Limitations] The limitation of this study is that the number of technical feature phrases extracted is not large.In the future,the patents and technology demands will be more comprehensively combined with keywords and technical feature phrases.