TF-IDF Based Contextual Post-Filtering Recommendation Algorithm in Complex Interactive Situations of Online to Offline: An Empirical Study

Abstract

O2O accelerates the integration of online and offline, promotes the upgrading of industrial structure and consumption pattern, meanwhile brings the information overload problem. This paper develops a post-context filtering recommendation algorithm based on TF-IDF, which improves the existing algorithms. Combined with contextual association probability and contextual universal importance, a contextual preference prediction model was constructed to adjust the initial score of the traditional recommendation combined with item category preference to generate the final result. The example of the catering industry shows that the proposed algorithm is more effective than the improved algorithm.

Keywords

• context information
• contextual post-filtering recommendation
• contextual preference
• item category preference
• TF-IDF