17 A Survey of Techniques for Developing Recommendation Systems

The amount information handled by information systems is growing steadily. This requires the use of recommendation systems for handling all this information. A recommendation system helps to find items of interest, but the development of this kind of systems is complicated due to the existence of a large number of algorithms for recommendation systems. In this context, this paper proposes an analysis of the algorithms used in recommendation systems and its most common application domain. The algorithms to calculate similarity presented in this paper are the algorithm of the cosine and Pearson correlation. The clustering algorithms analyzed are K-nearest neighbors (KNN) and clustering. All these algorithms are used in the collaborative filtering. The algorithms used for content-based filtering and hybrid approaches are Naive Bayes Classifier, Floyd-Warshall algorithm, Demographic recomm