Due to the ever increasing amount of information in the Internet, people will face various choices/options while accessing this information, which makes it very hard and even sometimes impossible to find their desired results. For example in music and video online services, users have access thousands of items. To address this issue, the recommender systems aim at providing intelligent recommendations to users, in order to help them search this wide range of available products and provide them with useful recommendation regarding users’ needs and preferences.
This part of our research activity is devoted to investigate the recommender systems. Here we study how to use many information sources to make better suggestions to users.
A novel context-aware model to improve quality of recommender systems
Recommender systems are among research fields in data mining and information retrieval, which due to their use for commercial purposes are widely used in many studies. The use of contextual information has recently received/gained more attention in this field. Contextual information contains the context and situation of a user which is interacting with the system at the time, such as the time or location of user while interacting, connection device or her mood. Each of these categories of information are used separately in different studies. Considering the fact that contextual information has received less attention in recommender systems, in contrast to their importance and value in information, studying the use of these features (specifically location information) can significantly improve the performance of recommender systems.
By analyzing the user’s location in her behavior, we aim to improve the quality of the recommendations of the new model compared to previous classic models which were independent of the location. The proposed model can have many applications in cyberspace and information systems; especially the applications in which options (e.g. products or services) available in the system have location labels. For example, the model can meet the needs of the recommender systems for entertainment and service website.
People involved: Ali Abbasi, Ali Khodadadi, Hamid R. Rabiee