A Framework for an Efficient Recommendation System Using Time and Fairness Constraint Based Web Usage Mining Technique

Users prefer to use various websites like Facebook, Gmail, and YouTube. We can make the system predict what pages we expect in future give users they have requested. Based on data gathered analyzed, user's navigation patterns response requests. In order track down users’ navigational sessions, web access logs created at a specific website are processed. Grouping user session is then done into clusters, where inter-cluster similarities minimized, although intra-cluster maximised. Recent clustering fairness analysis research has focused centric-based methods such as k-median k-means clustering. propose improved constrained based (ICBC) fair algorithms for managing Hierarchical Agglomerative Clustering (HAC) that apply constraints regardless of distance linking parameters, simplifying trials HAC intended protected groups compared vanilla techniques. Also, this ICBC used select an algorithm whose inherent bias matches problem, adjust optimization criterion any distinct take interpretation improve efficiency show our proposed finds fairer by evaluation NASA dataset balancing problem.