Constructing topic-based Twitter lists

[摘要] ENGLISH ABSTRACT: The amount of information that users of social networks consume on a dailybasis is steadily increasing. The resulting information overload is usuallyassociated with a loss of control over the management of information sources,leaving users feeling overwhelmed.To address this problem, social networks have introduced tools with whichusers can organise the people in their networks. However, these tools do notintegrate any automated processing. Twitter has lists that can be used toorganise people in the network into topic-based groups. This feature is apowerful organisation tool that has two main obstacles to widespread useradoption: the initial setup time and continual curation.In this thesis, we investigate the problem of constructing topic-based Twitterlists. We identify two subproblems, an unsupervised and supervised task,that need to be considered when tackling this problem. These subproblemscorrespond to a clustering and classification approach that we evaluate onTwitter data sets.The clustering approach is evaluated using multiple representation techniques,similarity measures and clustering algorithms. We show that it is possible to incorporate a Twitter user's social graph data into the clustering approachto find topic-based clusters. The classification approach is implemented,from a statistical relational learning perspective, with kLog. We show thatkLog can use a user's tweet content and social graph data to perform accuratetopic-based classification. We conclude that it is feasible to construct usefultopic-based Twitter lists with either approach.

[发布机构] Stellenbosch University