No abstract

Recommender systems offer a solution to the problem of successful information search in the knowledge reservoirs of the Internet by providing individualized recommendations. Content-based and Collaborative Filtering are usually applied to predict recommendations. A combination of the results of the above techniques is used in this work to construct a system that provides precise recommendations concerning movies. The content filtering part of the system is based on trained neural networks representing individual user preferences. Filtering results are combined using Boolean and fuzzy aggregation operators. The proposed hybrid system was tested on the MovieLens data yielding high accuracy predictions.

No abstract

Recommender systems provide a solution to the problem of successful information searching in the reservoirs of the Internet by providing individualized recommendations. Content-based filtering and collaborative filtering are usually applied to predict these recommendations. In this work a clustering approach based on semi-supervised learning is proposed. The method is then used to construct a recommender system for movies that combines contentbased and collaborative information. The proposed system was tested on the MovieLens data set, yielding recommendations of high accuracy.

No abstract