Serendipity through Social Search

The filter bubble of recommendation systems

The recommendation filter bubble – Eli Pariser

Aim: To stimulate serendipitous discovery of new web content by leveraging the Social Network Graph to augment search results.

 

The web provides us with a wealth of new information, knowledge and discussion on an incomprehensibly massive scale. The ability to stumble across fresh and interesting online content is facilitated by the likes of search tools, Social Networks and a number of highly visited websites. However each of these enablers of discovery are driven by some form of filter system in order to tailor potential content of interest to it’s users.

Recommendation systems play an increasing role in the content that we access. Each system attempting to feed personalised content to the user based on a number of factors and topics. Although these systems are able to provide a more context aware search experience, it becomes increasingly difficult to venture past pre-filtered recommended content and discover new content available out there on the web.

Therefore this project aims to provide alternative methods of discovering new thought provoking online content which would otherwise be missed by existing recommendation systems. By understanding ties of interest in users based upon their Social Network graph it may be possible to augment and generate search results which promote serendipitous discovery of new online content.