Ranking tweets as search results for a query is challenging because of the massive number of tweets that are being generated in real time as well as the short length of each individual tweet. Recently we explored re-ranking of twitter search results. The algorithms explored earlier have significant scope for improvement to get more efficiently ranked search results. In this thesis we will elaborate on the re-ranking functions based on the content and social net- work properties and also include the temporal dimension to this investigation to enhance search results and improve its utility. Moreover, we will improve the underlying data collection framework and publish these datasets online for review and use by other researchers working in this domain.
Re-ranking of twitter search results
Institute of Technology