Advances in Clustering Collaborative Filtering by means of Fuzzy C-means and trust
Expert Systems with Applications: An International Journal
Web 2.0 Recommendation service by multi-collaborative filtering trust network algorithm
Information Systems Frontiers
| Hi-index | 0.00 |
Recommender Systems (RS) depend on users' previous opinions and other users' opinions on items suggest to them items they will like. However, other users' opinions are often unreliable, such as malevolent remarks. In order to reduce the influence of malevolent remarks and improve accuracy of recommendation, the authors propose a Trust-Aware Recommender Model (TARM), which can utilize trustworthy experts and their search experiences to recommend their search histories to the common user according to profile similarity between common user and experts. In addition, the authors also discuss the core of this model -- an algorithm to compute profile similarity in a community. And in order to illuminate and validate this method, the authors have implemented the above model and algorithm through extending the open source search engine "Nutch".