The Design and Implementation of an Intelligent Online Recommender System

Abstract

Recommender systems(RS) are intelligent applications in the field of information retrieval. Information retrieval assists users to take part in decision making process. It assists them in choosing one item from a vast set of alternative products or services. The scope of recommender systems has expanded gradually over the time from 1990s. As the user provides the inputs, these inputs are recorded and used as recommendations. These inputs are further used in the recommender system tool. The inputs received by the user are aggregated by the other people’s inputs and then the system sends them directly to the appropriate recipients(Dean etal.,1995). RS is basically a technology which is based on the important aspects such as collaborative or social filtering. There are many researches already in the area for collaborative filtering and social filtering(Bigus,1996; Rennieand Srebro,2005). The RS can be used in intelligent information retrieval in the field of artificial intelligence. From information retrieval, recommendation technology explores and derives the vision that users are searching. In the process of recommendations, the user is engaged in information searching and the RS automates and collects the content required for matching purposes. Typically the search results are arranged in the form of a ranked list. One of the important phases of artificial intelligence is learning process. This will view the past knowledge, buying behaviour, and interest. The RS are primarily based on two phases that are search phase and user-based interaction model. These phases can be identified as user-model construction and recommendation generation. This paper defines the importance of both the models. The interaction model describes the user needs and preferences. Based on the interactions, during the session they are connected to the same interest group. This paper attempts to define a proposed model for considering the factors of intelligent retrieval.

Keywords: Recommender Systems, Intelligent Applications, Retrieval, Filtration

View PDF