A Heuristic Method to Recommendation Systems

Abstract

Recommendation systems are a subclass of information filtering systems that seek to predict which items are most likely interesting to the user. Content-based recommendation systems recommend items that are similar to the previously liked items. Context-aware recommendation systems are a type of content-based recommendation systems, where the items are filtered based on the value of their attributes. In this paper, a general mathematical model was developed for context-aware recommendation systems using a heuristic method. This method was implemented as a multiuser demo application, a movie recommendation system. The application was tested and the results have been evaluated with different metrics. It could be declared that the developed method is more precise than the existing methods.