Menu

Exploring collaborative annotations in hybrid recommendation systems

Faculty Involved: Marcelo Manzato

Recommendation services are an essential tool for dealing with information overload. However, a common problem that exists is the knowledge of meaningful information about content and user preferences. The difficulty of obtaining this information is called the semantic gap, and researchers over the years have studied related issues. On the other hand, with the advent of Web 2.0 and the ability for users to act as content producers and increment data with annotations, new search possibilities were created to reduce the effects of the semantic gap. This research plan aims to investigate some of the challenges related to using collaborative annotations to improve referral services. To this end, it is proposed to develop a unified recommendation model capable of analyzing the information produced by the interaction of users with the system, automatically obtaining richer metadata about the content, as well as about the personal interests of individuals. The expected results of the project include the efficient integration of techniques from different areas, such as information retrieval, machine learning, and natural language processing, in the context of recommendation systems.

Translate »