Recommendation with User Trust and Item Rating

Abstract

Recommender systems is becomes widespread and utilized in several fields for gathering the knowledge supported the user necessities. It�s in the main wont to facilitate the user for accessing the method supported the relevant data. Several framework for recommendation systems supported the various algorithms area unit revolve round the idea of accuracy solely however alternative necessary feature like diversity of the recommendations area unit neglected. The main idea of these works is that not only incorporating demographic information of users in profile matching process of CF-based algorithms is important weighting should be assigned to these features including rating feature the motivation behind this idea is that �different users place different importance or priority on each feature of the user � profile. For example if a male user prefers to be given recommendations based on the opinions of the other men then his feature weight for gender would be higher than other features�. Here we apply improved invasive weed optimization (IIWO) algorithm for the same purpose with some little changes in selecting the potential similar users as described in the previous sub section and in the evaluation criteria. After the optima weights have been found the two profiles are compared according to equation based on the Euclidean distance of the two profiles.