Recommendation system  is the type of information filtering which involves prediction of rating and user preferences, which would help user to buy items according to their needs and interest. The books suggestion at Amazon.com is best example of recommendation system .Recommendation System directs  the way to find products, Information according to their interest. Recommendation system uses following technologies to recommend products: Content filtering, Collaborative filtering & Association mining. Content filtering recommends item which is based on Users profile, which he has liked in past. Collaborative based filtering is method to analyze the user’s behavior by predicting the users taste to that of similar to other user. Association mining serves as finding relations/correlation among items in a large database. An association rule is a condition of the form X ĺ Y where X and Y are two sets of items. It means it finds correlation between X and Y i.e. If we buy X item then it finds chances to buy Y items.
Recommendation systems is used for the purpose of suggesting items to purchase or to see. They direct users towards those items which can meet their needs through cutting down large database of Information. A various techniques have been introduced for recommending items i.e. content, collaborative and association mining techniques are used. This paper solves the problem of data sparsity problem by combining the collaborative-based filtering and association rule mining to achieve better performance. The results obtained are demonstrated and the proposed recommendation algorithms perform better and solve the challenges such as data sparsity and scalability.
Today the huge amount of information is available online due to the acceptance and understanding of the possibilities of internet. This reason makes the World Wide Web as an important research area. Sarwar, et al.,, has introduced and analyzed effect of different similarity algorithms and showed the experimental results through the prediction h MAE graph and also proposed that size o neighborhood effects prediction quality. Hongwu ye  proposed a method for finding nearest neighbor through self-organizing map which makes a group of nearest neighbors which is first step in collaborative filtering. Association mining is used to fill vacant space .Thus they proposed combination of association mining and SOM to address the issue of data sparsity. Hengsong Tan, et.al;  presented a new approach to address the issue of data sparsity problem by combining item classification and item based collaborative approach.
Existingrecommender systems since the existing only considers the recommending the items based on user ratings of item. It doesn’t recommend items when ratings for an item are not available.
The proposed system uses combination of collaborative filtering and association mining. Collaborative filtering is used for finding similarity between items which would help the system to recommend items and association mining is used for filling the vacant ratings where necessary. Then it uses prediction of target user to the target item using item based collaborative filtering.
Thus the use of both methods can help to manage data sparsity problem and cold start problem in recommender system.
- Operating system : Windows.
- Coding Language : Python.
System : Pentium IV 2.4 GHz or intel
Hard Disk : 40 GB.
Floppy Drive : 1.44 Mb.
Mouse : Optical Mouse.
Ram : 512 Mb.
The increasing demands of Online Information have lead to invent new techniques for prioritizing and presenting items of Users Interests. This paper uses item-based Collaborative Filtering. To produce ratings .The Item based collaborative filtering can remove the data sparsity problem and can provide good recommendation. Finally the results of similarity calculation give good performance at accuracy.