Research Status and Future Optimization Direction of Domestic and Foreign Patent Recommendation Technology Based on Patent Data Mining
The purpose of this paper is to deeply excavate the feature information and correlation relationship embedded in patent data,comprehensively sort out and analyse the current research status of patent recommendation technology,and systematically summarise the key points and difficulties of the research as well as the future research direction,so as to promote the optimization of patent recommendation system.By adopting the literature research method and content analysis method,this paper outlines the relevant researches on patent feature selection and recommendation technology methods from the beginning of the 20th century to the present,and analyses their stage-by-stage development.In view of the current research status of patent recommendation,it summarises and reviews the shortcomings of the existing research and looks forward to the future optimization direction.At present,most of the feature selection in patent recommendation technology stays in explicit features and fails to fully explore the implicit features in the data.In addition,although the existing research has experienced the transformation from single graph structure,deep learning model to the combination of the two,there are still some deficiencies in terms of interpretability.The future optimization direction should be devoted to improving the interpretability of deep learning models and deeply mining the implicit feature information in the data.
万方数据
维普
