Abstract:Collaborative filtering over the years have emerged as an alternative recommender system to address some of the setbacks of content based filtering. Although, Collaborative filtering has offered some benefits to the majority of the online stores in recommending products to users using users' ratings of similarity measure, its usage has also raised some doubt in the minds of researchers, regarding its effectiveness in handling ratings with limited number of users or no rating record from users. Thus, this has resulted in efforts by researchers in determining further ways to combat the issues attributed with the existing collaborative filtering techniques in terms of data sparsity or cold-start situations. This study focused on improving the traditional similarity measurements that currently exist on the item-based collaborative filtering, in order to accommodate and mitigate further the issue of cold-start situations. Thus, this study proposed an algorithm which is meant to balance the three current traditional measurement metrics such as: Cosine-based similarity, Pearson correlation similarity and Adjusted cosine similarity, in the direction of cold-start situations. The improved algorithm of traditional measurement metrics were further compared with the existing algorithm of the traditional metrics. Results showed that the proposed algorithm offered a better item-based collaborative filtering algorithm to recommendation systems than the existing, using data set from Movielens recommender system. Hence, the proposed algorithm did not only mitigate the drawbacks experienced with the three traditional algorithms in terms of data sparsity or cold-start situations but also retained the good features of the existing item-based collaborative filtering algorithm. Thus, the proposed algorithm complemented the strength of the three traditional measurement metrics with evidence shown when measured with Mean Absolute Error.

Improving Jaccard Index for Measuring Similarity in Collaborative Filtering

Employing Group Interest Preference Degree in Collaborative Filtering Recommendation

Improving Recommendations With Collaborative Factors

An improved collaborative filtering method based on similarity

A Combinative Similarity Computing Measure for Collaborative Filtering

Nonparametric statistics inspired similarity measure for accuracy in collaborative filtering

Collaborative Filtering Recommendation Algorithm based on Improved Similarity

A new item similarity based on α -divergence for collaborative filtering in sparse data

Empirical comparison of local structural similarity indices for collaborative-filtering-based recommender systems

An effective collaborative filtering algorithm based on user preference clustering

Collaborative filtering recommendation algorithm based on multi-level hybrid similarity

A two-step Recommendation Algorithm via Iterative Local Least Squares

An Integrated Method for Similarity Computation

A New User Similarity Model to Improve the Accuracy of Collaborative Filtering

A Collaborative Filtering Algorithm for Sparse Data.

Research on Collaborative Filtering Recommendation Algorithm for Improving User Similarity Calculation

Alleviating the Sparsity Problem of Collaborative Filtering Using Rough Set

A New Similarity Computation Method in Collaborative Filtering Based Recommendation System.

An Analytical Approach to the Jaccard Similarity Index

A New Similarity Model Based on Collaborative Filtering for New User Cold Start Recommendation

An optimized item-based collaborative filtering algorithm