My bachelor project has the title “Automatic Video Recommendations”. I don’t feel like writing all over again about it in here after the 58 pages written in the thesis but I like the video of the developed application so here it is.
The objective of this project is to create a system that merges social and videos data to recommend videos of interest to the user using both the Item based and the collaborative approach. That should be done with the least possible effort from the users side. That is, we will use a social website, Twitter, to get an idea of what are the users interests and use the information to recommend videos from YouTube in addition to matching the user with others who have similar interests. Not only that the user does not have to enter his personal information all over again, but also by providing a web-browser plugin, the user does not have to visit a certain page or website to get the recommendations. Our aim is also to provide users who use English, Arabic and Franco-Arabic in their online profiles with appropriate recommendations.
Technically, this should be done by mining the videos available information like title, description and comments in addition to the users tweets, analyze them and build a profile for the videos and the users. After that, the users are matched with the videos and users are matched with other users to provide both Item-based and Social recommendations.
For a video demo, please click here.