Quality Of Similarity Measure Research Articles

A personalised recommendation algorithm based on probabilistic neural networks

Collaborative filtering is widely used in recommendation system. Our work is motivated by the observation that users caught in their attention relationship network, and their opinions about items will be directly or indirectly affected by others through such a network. Based on behaviours of users with similar interest, the technique focuses on the use of their opinions to recommend items. Therefore, the quality of similarity measure between users or items has a great impact on the accuracy of recommendation. This paper proposes a new recommendation algorithm with graph-based model. The similarity between two users (or two items) is computed from the connections on graph with nodes of users and items. The computed similarity measure is based on probabilistic neural networks to generate predictions. The model is evaluated on a recommendation task which suggests that which videos users should watch based on what they watched in the past. Our experimental results on the YouKu and Epinions datasets demonstrate the effectiveness of the presented approach in comparison with both collaborative filtering with traditional similarity measures and simplex graph-based methods and further improve user satisfaction, our approach can better improve the overall recommendation performance in precision, recall and coverage.

A personalised recommendation algorithm based on probabilistic neural networks

Collaborative filtering is widely used in recommendation system. Our work is motivated by the observation that users caught in their attention relationship network, and their opinions about items will be directly or indirectly affected by others through such a network. Based on behaviours of users with similar interest, the technique focuses on the use of their opinions to recommend items. Therefore, the quality of similarity measure between users or items has a great impact on the accuracy of recommendation. This paper proposes a new recommendation algorithm with graph-based model. The similarity between two users (or two items) is computed from the connections on graph with nodes of users and items. The computed similarity measure is based on probabilistic neural networks to generate predictions. The model is evaluated on a recommendation task which suggests that which videos users should watch based on what they watched in the past. Our experimental results on the YouKu and Epinions datasets demonstrate the effectiveness of the presented approach in comparison with both collaborative filtering with traditional similarity measures and simplex graph-based methods and further improve user satisfaction, our approach can better improve the overall recommendation performance in precision, recall and coverage.

Lọc cộng tác với độ đo tương tự dựa trên đồ thị

Collaborative filtering is a technique widely used in recommender systems. Based on the behaviors of users with similar taste, the technique can predict and recommend products the current user is likely interested in, thus alleviates the information overload problem for Internet users. The most popular collaborative filtering approach is based on the similarity between users, or between products. The quality of similarity measure, therefore, has a large impact on the recommendation accuracy. In this paper, we propose a new similarity measure based on graph models. The similarity between two users (or symmetrically, two products) is computed from connections on a graph with vertices beeing users and products. The computed similarity measure is then used with k – nearest neighbor algorithm to generate predictions. Empirical results on real movie datasets show that the proposed method significantly outperforms both collaborative filtering with traditional similarity measures and pure graph-based collaborative filtering.

A COLOR IMAGE RETRIEVAL METHOD BASED ON COLOR MOMENT AND COLOR VARIANCE OF ADJACENT PIXELS

This paper first introduces three simple and effective image features — the color moment (CM), the color variance of adjacent pixels (CVAP) and CM–CVAP. The CM feature delineates the color-spatial information of images, and the CVAP feature describes the color variance of pixels in an image. However, these two features can only characterize the content of images in different ways. This paper hence provides another feature CM–CVAP, which combines both, to raise the quality of similarity measure. The experimental results show that the image retrieval method based on the CM–CVAP feature gives quite an impressive performance.