ABSTRACT
Recently Recommender Systems is widespread in WWW in which user can share their opinions and experiences by ratings or descripting figure of items. The consumer can spontaneously bookmark an item in social bookmarking by giving various descriptions of an item. Even though social networking sites by giving a description of an item, the system can predict whether it is a positive or negative opinion of that particular item and it can be recommended to other users who have the similar interest in that. Generally, RS is classified into content-based, in which the user preferences of ratings are based on his/her previous history of his/her likings. Collaborative Filtering, recommendations are done based on consumer’s relationship. The major research work done in Collaborative filtering because as data is increasing the prediction of advice is additionally tough to seek out therefore there would be a bunch of users having similar taste and Hybrid Filtering, is merging technique of Content-based filtering and Cooperative Filtering. Here the three techniques are merged so that offline and online information are used efficiently to avoid scalability and sparsity problems in recommendation systems. Adding Real Value Genetic algorithm as a special feature to the problem where the density of users are more in datasets and samples are discretized where the variables are continuous. This paper conjointly analyzed that recommendations are influenced by the factors like age, gender, occupation and a few alternative user profile data. This work consist both content and collaborative Filtering methods and certain demographic information are merged into a hybrid approach, wherever further content options are used to enhance the exactness of Recommendation Systems. Conjointly, it tends to use the Real value genetic algorithm to get accurate results and to provide recommendations to the user.
