Science & Engineering Research Support soCiety (SERSC)
Collaborative filtering is of particular interest because its recommendations are based on the preferences of similar users. This allows them to overcome several key limitations. This paper explains the need for collaborative filtering, its benefits and related challenges. The authors have investigated several variations and their performance under a variety of circumstances. They also explored the implications of these results when weighing K nearest neighbor algorithm for implementation. Based on the relationship of individuals, putting forward a new incremental learning collaborative filtering recommendation system, discovery it is a better way to acquire optimum results.