TY - JOUR T1 - Linear Time Growth Collaborative Filtering Based on Caching Techniques AU - Waleed M. Al-Adrousy , Hesham A. Ali, and Taher T. Hamza JO - Journal of Information and Computing Science VL - 3 SP - 181 EP - 188 PY - 2024 DA - 2024/01 SN - 9 DO - http://doi.org/ UR - https://global-sci.org/intro/article_detail/jics/22578.html KW - Social networks, collaborative filtering, recommender systems, caching techniques. AB - One of social networks features is to recommend some items for users suitable and their profiles and needs. One of the used techniques of item recommendation is to calculate the average rating of other users' ratings of that item. The calculated average can be used as a rank for that item. New users' ratings are made within the lifetime of social networks. This dynamic sized set of ratings can lead to a performance problem. Ranking is not fixed since new users are constantly evaluating items in networks. However, updating ranking by traditional averaging can have a quadratic time growth. This paper aims to convert this aggregation method into linear growth instead of quadratic. The suggested algorithm needed extra storage capacity. Caching is suggested to speedup ranking with preserving storage resources. Another target is to have a high hit-ratio with better utilization of small cache size compared to the traditional Greedy- Dual- Size-Frequency algorithm. At the end of this paper an evaluation of the proposed technique performance is provided based on simulated experimental results. The experiment results show that the proposed technique has better performance rather than traditional averaging recommender systems without caching.

Waleed M. Al-Adrousy , Hesham A. Ali, and Taher T. Hamza. (2024). Linear Time Growth Collaborative Filtering Based on Caching Techniques. Journal of Information and Computing Science. 9 (3). 181-188. doi: