Talash

= Finding People You Already Know =

StatusNet Google Summer Of Code 2010

= Introduction =

This task will allow users to break their boundaries, make new friends, find existing friends across StatusNet powered domains and follow, interact and expand their thoughts with them. The approach will tap into a users contact list on his other public profiles. Search existing friends with OStatus accounts using the Social Graph API and FOAF. Mark possibly new friends and suggest the user to follow them.

= Deliverables =

The task will be tackled on various frontiers by trying to gather information regarding the users existing friends network on domains like twitter or facebook. A provision to upload ones contact list or extract from their webmail accounts to have a strong list of people known to the user will created. These profiles can be searched for on the same network for a quick collection of people known to the user. Social graph API will be used to dig out a users public connections. FOAF will also be used as base for searching friends with similar interests. This re-enforced with the users contact list will be used to develop new connection requests. The goal is to allow a new user to quickly find familiar people and existing users to expand their network for better inter-networking.

The approach for searching friends across federated networks is two fold.

 == Quick Connect ==

 Using PoCo to tap into google accounts of users.  Using API&rsquo;s to retrieve contacts list of their facebook/twitter profiles. The email address, name and other information can be used to dig into the Social graph of users public profiles and return a list of friends that might be known to the user. This list of suggestions will be limited to one hop distance from the user. This mean that only those people who are directly connected to the user in their other public social existences will be returned. This is to allow a quick connect to familiar faces.   == Delayed Conenct ==

The motivation behind this approach is that a new user, a reserved user who is choosy about his friends, or other users who want to extend their network for business or similar interests must not be bombarded with a huge of potential friends who can be subscribed to. This is common place in many other Social networking websites. It confuses the user and creates links which might be never used i future. Social graph API inherently uses FOAF. However the this approach will use FOAF of a user on its own to detect potential friends on other networks. The idea is to involve interests, geographical proximity and interconnections on other public profile pages of users, give them scores of positive value, and assign them as weights on the social graph of the individual. A simple heuristic may be used to traverse the graph one step at a time and find friends which can be potential friends in future with high probability. 