An Improved Recognition Method of Weighted Rules and Its Application in Recommendation Algorithm

2022 Academic Journal of Computing & Information Science  
The method proposed in this paper mainly uses the difference of the importance of items in the database to improve the problem of mining a large number of redundant and useless rules by traditional association rules mining algorithm. The method has made great improvement in three aspects of association rule design, which are effective length recognition of recommendation rule, weighted association rule mining combined with frequency analysis and time constraint application. The weighted
more » ... association rules mined by the effective frequency length weighted association rules mining algorithm can improve the accuracy of recommendation, and the accuracy of recommendation prediction is increased from 62% to 69%. The purpose of this algorithm is to mine hidden high value rules, and optimize the algorithm considering the time complexity of algorithm execution while ensuring the accuracy of the algorithm.
doi:10.25236/ajcis.2022.050102 fatcat:gmrhlrbu5fhb5hh457bgeii3qu