A Comparison between Intra-Transaction Association, Inter-Transaction Association and Collaborative Filtering

Using the techniques of Collaborative filtering based on historical records of items purchased by users, Recommender systems suggest items to users. Recommender systems use techniques of data mining for determining the similarity among a collection of data items, by analyzing user data and finding hidden useful information or patterns. The Collaborative filtering technique tries to find relationships between the existing data and new data and determines further the similarity and provide recommendations. In this paper, the intra-transaction association, inter-transaction association and collaborative filtering approaches are compared. The similarity between companies is compared in different cases using collaborative filtering technique and accordingly recommendations are generated for users interested in investing in stock market.