Method for Improving the Similarity Measure of Sparse Scoring Based on the Bhattacharyya Measure
DEStech Transactions on Computer Science and Engineering2017Iss. aita
Abstract
The similarity computation has an important influence on the performance of the personalized recommendation system. Due to the sparsity of the data in the recommendation system, the traditional similarity measurement method can not achieve the desired results. Sparse scoring similarity measurement method based on the Bhattacharyya measure has achieved good results. In this paper, the score similarity measure is introduced. Then, a method to improve the similarity measure based on the Bhattacharyya measure is presented. Taking into account all the items of the two users’ score, and according to the score to distinguish between the user is really interested in the project.
Related Papers
- → An improved item-based collaborative filtering using a modified Bhattacharyya coefficient and user–user similarity as weight(2022)34 cited
- → An Improved Spatiogram Similarity Measure for Object Tracking(2018)3 cited
- → Similarity measure design for high dimensional data(2014)1 cited
- → Quantisation error curve: A measure of similaritybetween requantised images(1994)
- → Method for Improving the Similarity Measure of Sparse Scoring Based on the Bhattacharyya Measure(2017)