Mobility-assisted SocialNetby Wireless Mesh Network Group, N
In this project, we focus on applying the location information and mobility patterns of users to facilitate their social networking communications in providing location-based services.
The basic idea of this project is to analyze the location distribution and mobility trajectories of cellphone users to study their routines and patterns, and employ such information to design communication protocols for mobile social networks. For example, people with similar mobility trajectories and routines may have similar interest in their information access and location-dependent services, which thus helps to enhance the quality of friend recommendation and location-based service provision in mobile social networking.
Firstly, we use data mining techniques to discover mobility patterns from users' movement datasets. To be more specific, we focus on the temporal and geographical regularity of user's movement. Based on the observation that human beings' movement is periodical in time dimension and has distinguishing characteristics in spatial dimension, we can optimize mobile local communications. Secondly, we pay attention to the co-location of different users based on their movement data. Several problems is proposed on this issue, such as how likely are people going to be in the same location, what's the distribution of people's co-location relationships and what are the factors influence the co-location relationships. Finally, we apply the research results about mobility and co-locations to facilitate mobile social networking communications.
The collected user data is only used for academic research. We will strictly follow the privacy regulations indicated in the website and the privacy laws in the country. The collected data will be anonymized before analysing, where users identities are replaced by random numbers using multiple hashing functions. Users personal information such as cellphone numbers, cellphone records will not be recorded or will be deleted immediately after data collection. The dataset is only access-able to the researchers in the university and it will not be published in public.
Firstly, we'll publish paper on what we learn about the users' movement and how we impose them on local communication platform.
Then a local social network service, Gossipdog, will be provided in the future.