A User Timeline Summarization Technique using TextRank Algorithm
In-Seok An, Hyun-Woo Kim, Hyoung-Joo Kim · 2012
As digital device has come into wide use, information streams have recently emerged as a popular means of information awareness. Twitter is one of the most popular micro-blogging service and social media with a limit of 140 characters. If a user follows Twitter accounts continuously, the user may subscribe information more than the user can process. In this paper, we apply TextRank algorithm to alleviate information overload in Twitter user's timeline. After modeling user timeline as a graph, we apply graph-based ranking algorithm to the graph of timeline. Based on the score of each vertex, we apply concept of summit to summarizing user timeline. The experimental results show that proposed method summarizes user timeline more effectively than existing method that rely mainly on frequency based method.