People to people recommendation deals with the problem of finding meaningful relationships among people or organisations. This research aims at initiating the theoretical development and system implementation on people to people recommendation.
In online social networks, relationships can be friends on Facebook, professional contacts on LinkedIn, dates on an online dating site, jobs or workers on employment websites, or people to follow on Twitter. The nature of these domains makes people-to-people recommender systems to be significantly different from traditional item-to-people recommenders. One basic difference in the people recommender domain is the benefit or requirement of reciprocal relationships. Another difference between these domains is that people recommenders are likely to have rich user profiles available. This research will enable people and organisation to find other people or organisations easily which will consequently lead to an improvement on user online experience and an increase in productivity.
This research has been funded by a Smart Services CRC project. We conduct the following key research investigations in the scope of this project:
(1) Data analysis of people to people interactions in real world social networks
(2) Effective and efficient general recommendation methods
(3) Effective and efficient methods specified to people to people recommendation problems
(4) Building people to people recommender systems
We have finished majority of the research now. We are currently working on implementation of the developed people to people recommender system for potential deployment in an real world commercial social network.
We have developed several people to people recommendation methods based on user behaviours or profiles as well as hybrid methods with excellent performance. We have built prototype recommender systems that supports people to people recommendation on historical real world data and online real world commercial systems. Both prototype systems and trials have highly performance and support very large scale data. Surprising findings derived from our results show that user actually behaviours are often different from users' explicit preferences in terms of relationship development.
With such results, it is now possible for companies to provide better personalisation to their users, and for users to have better online experience by using personalised and recommended information.