Query construction based on concept importance for effective patent retrieval

Feng Wang,Lanfen Lin

DOI: https://doi.org/10.1109/fskd.2015.7382158

2015-08-01

Abstract:Patent retrieval is a long query task whose aim is to retrieve all documents related to patent applications. However, current approaches face with the term mismatch problem, leading to low retrieval performance. To deal with this issue, we propose a novel automatic query construction approach based on semantic concept importance for effective patent retrieval. In this approach, natural language processing techniques are firstly adopted to analyze patent long query inputs. Then, candidate query concepts are generated according to the concept features. Further, a concept importance-based query construction algorithm is presented to select the representative query concepts. Experimental results on the standard patent dataset demonstrate that our proposed approach can significantly outperform other state-of-art methods.