In today’s information society, navigating the lush of the enormous amounts of data surrounding us is a challenge that is very commonplace, although hard to automate. The massive flow of data from social networks, news outlets and daily social communication – be it in the form of physical of digital – does not give us an easy way to extract or prioritize relevant information.
HintMe is an intelligent recommendation engine and algorithm, which is able to learn from the user’s preferences to suggest content on the web. The engine, which is based on system using machine learning, can give us ever more precise and relevant suggestions and recommendations for the users when the number of the users and ratings increase. HintMe learns from the individual’s preferences and matches those preferences to other users’ habits.
In my thesis I outline some of the concerns about today’s most relevant recommender systems and algorithms. Exploiting the weaknesses of these systems I present to the Reader a new method, which fixes the aforementioned weaknesses. I guide you through HintMe’s basic algorithm, the system’s architecture, implementation, looking at potential problems and further development scenarios. I visualize a few of the many possible applications of the HintMe system starting with the obvious advertisement targeting, along with the possibility of social experience applications all the way to a system that’s able to deduce long term business deductions based on the system’s users’ explicit habits.