Web 2.0 has provided a new way for people to communicate with each other. Nowadays, Users are often faced with an information overload. Even in the days before Internet, people found it difficult to decide what book to read, or what movie to watch. They were often guided with the opinions and recommendations of their friends. There are also some situations, when users do not know exactly what they want or they may not even be actively looking for information. Generally, users have some particular interests and they will like to have WebPages, news articles, blogs or events related to their interests delivered to them. However, they are currently forced to visit multiple sources and scan through irrelevant content before finding useful information. For such a long-term information need, one of the best ways to help users is to recommend information to them.

With a number of social networking Web sites such as Facebook, MySpace, Orkut and LinkedIn available, it is most desirable to have a system application that could integrate information from multiple sources to provide customized information for a community, i.e. a group of users sharing some common interests. These social communities allow us to get feedback from users and develop algorithms that most users would benefit from.

In this project, we propose to build a news recommender system for popular social network, Facebook. The system prepares daily news letters for communities on Facebook. The users register a community by providing a keyword description and a set of news sources. The system then fetches the news articles and then filters them based on the community description to prepare daily news digest.

Fig. 1: Snapshot of the System

An online demo is available at here In our system, we support:

• Registering a New Community. A user can submit a keyword descriptions and set of news sources. The system would then fetches the news results and filters them based on the community descriptions.
• Browsing a list of Communities. A user can browse the list of registered community by clicking on the "Browsing tab". Clicking on each of the communities will show the daily news articles filtered based on the description of the community.
• Rating a Document. A user can rate a document. The recommender system has a feature called "Popular articles" for each community. Every community has a list of articles that were most popular among its users in the last week. This list is prepared based solely on users' feedback.
• Recommending a Document. A user can recommend a news article to his friends on Facebook. When a friend makes a recommendation, his friend is sent a notification.
• Inviting friends to use the news recommender system.

Our main contribution is building a novel news recommender system and integrating it with Facebook and gathering user feedback.