A Weighted Slope One Collaborative Filtering Algorithm for Improved User Relevance
- Resource Type
- Conference
- Authors
- Li, Ruizhi; Liu, Qicheng
- Source
- 2023 International Conference on Artificial Intelligence and Automation Control (AIAC) AIAC Artificial Intelligence and Automation Control (AIAC), 2023 International Conference on. :1-4 Nov, 2023
- Subject
- Computing and Processing
Correlation coefficient
Automation
Collaborative filtering
Tagging
Filtering algorithms
Prediction algorithms
Information filters
slope one algorithm
data sparsity
collaborative filtering
matrix fill
personalized recommendations
- Language
Considering the low accuracy of recommendation results of current collaborative filtering algorithms caused by data sparsity, this paper proposes a weighted Slope One collaborative filtering algorithm for improved user relevance (WSOCF-IUR). The proposed algorithm takes into account the influence of the number of different user comments and the number of common rating items on recommendation accuracy. In addition, by incorporating user activity and improving the Pearson correlation coefficient, the algorithm reduces the influence of interfering data on recommendations and improving the overall recommendation accuracy. Experimental results demonstrate that the proposed algorithm achieves better recommendation results than other collaborative filtering recommendation algorithms.