Abstract: Recommender systems are now evolving as essential marketing and decision supporting tools in online advertising environment, because many customers who deal with e-commerce and related domains are often overloaded with complex and raw data driven by dynamic workflow operations, processes and business guidelines. Recommender systems will analyze and filter complex data methodically to provide essential and useful information to customers (in e-commerce). In this paper, we have presented an efficient recommender algorithm based on user browsing behavior. This algorithm will study user’s behavior, interests and other parameters to present them relevant ad(s) when they navigate through web pages. The Ad recommendation system works dynamically using input details about an advertisement and user past website visitation behavior. These inputs determine advertisement placement and presentation. The algorithm solves the problem of “recommending more related ad(s)” providing opportunity to generate higher user response, user satisfaction deriving more value for time and money. Our experiments and results prove that the dynamically recommended Ads by recommender system have got the relevance up to 92.47% out of 1900 instances.

Keywords: Recommender system, user behavior, online advertising.