International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395 -0056
Volume: 04 Issue: 01 | Jan -2017
p-ISSN: 2395-0072
www.irjet.net
Cross Domain Recommender System using Machine Learning and Transferable Knowledge Pooja Rawade, Shrushti Dhawale, Sudarshan Jagadale, Kunal Patil, Prof.Shweta Koparde 1Pooja
Rawade, 2Shrushti Dhawale, 3Sudarshan Jagadale, 4Kunal Patil
Student, Dept. of Computer Engineering, Pimpri Chinchwad College Of Engineering, Maharashtra, India 5Prof.
Shweta Koparde
Professor, Dept. of Computer Engineering, Pimpri Chinchwad College Of Engineering, Maharashtra, India
---------------------------------------------------------------------***--------------------------------------------------------------------because of the abundance of practical applications Abstract - The useful knowledge from an auxiliary domain can be transferred through the social domain to a that help users to deal with information overload and target domain. However, sometimes we may suffer from provide personalized recommendations, content and item cold start problem in the target domain. To alleviate services to them.[3] this issue we apply cross domain algorithm along with page ranking algorithm. The cross domain algorithm is divided into two stages; In the first stage we apply the TrAdaBoost algorithm to select some items which are being recommended to users in the target domain. Whereas, in the second stage we adopt nonparametric pairwise clustering algorithm to make a decision whether to recommend an item to user or not. The algorithm finds the recommended or not recommended customer groups for one item through the two stages and then with the help of page ranking algorithm we provide relevant and unsearched data to the users.
Key Words: Collaborative Filtering, Cross Domain, NonParametric Pairwise TrAdaBoost Algorithm.
Clustering
Algorithm,
1.INTRODUCTION
Recommender systems have become an important research area since the appearance of the first papers on collaborative filtering in the mid-1990's.[4][5][6] There has been much work done both in the industry and academia on developing new approaches to the recommender systems over the last decade. The interest in this area still remains high because it constitutes a problem-rich research area and Š 2017, IRJET
|
Impact Factor value: 5.181
|
A social networking service is a platform on which users can create and adopt different types of items such as messages, data or images. This huge volume of items generates a problem of information overload.[2] Thus with the development of information technology, we have already entered the era of big data. However, to discover the efficient data in various domains is critical.[7] Most recommender systems [8] encounter cold start problem. Cold start problem not only refers to a new user without any experience but also a new entity with few ratings and entirely a whole system. Cold start problem is challenging because no prior knowledge can be used in recommendation. Collaborative filtering is used to alleviate data cold start problem.[1] Another simple way is to transfer one domains information to the target domain i.e. cross domain recommendation. The data distribution (users, items & their features) in each domain is quite different and the new items in the target domain often suffer from cold start problem.
ISO 9001:2008 Certified Journal
|
Page 462