Title:
Social Network Discovery based on User Location Analysis

Description:
Today, an increasing number of people use Online Social Networks (OSNs), which primarily support their users in finding people, staying in contact with distant acquaintances or gathering information. Furthermore, sharing content, organizing events or social gaming are additional kinds of applications of OSNs.
Besides the explicit data acquisition, which is used by current OSNs, there is also data that can be collected implicitly. Location data is one type of information, which today can be easily obtained implicitly via GPS-enabled mobile devices. Current OSN services use this data to let contacts know about a user's whereabouts and to recommend Points of Interests (POIs).
In this thesis, implicitly collected location data should be used to explore the spatial social context of a person. The spatial social context comprises, for example, which people were near a specific user at a given time. Based on the spatial social context metrics to measure the spatial similarity between users have to be defined. Finally, a social network can be constructed, which is based on the users' implicitly collected location data and pictures spatial similarity between them. The resulting social network should be evaluated in one appropriate scenario in the form of a prototype.

Challenges:

• Definition and evaluation of several metrics for modelling the "location graph"
• Definition and implementation of suitable metrics in order to define the spatial similarity
• Evaluation of the implemented algorithms in a suitable application scenario

Prerequisites:

• Very good programming skills
• Interest in working in the research area of Social Networks

Supervisor: Prof. Dr. Axel Küpper [1], Sebastian Göndör [2], Peter Ruppel [3]

Type: Master Thesis

Duration: 6 months