Bilkent University
Department of Computer Engineering


Topic-Based Influence Computation in Social Networks


Kaan Bingöl
MSc Student
Computer Engineering Department
Bilkent University

A popular application of social network analysis is influential user computation and maintenance. Methods to analyze influence are important in understanding social trends. In our ongoing work, we study topic-sensitive influence computation using weights determined by a semantic analysis of social media posts and their spread in the network. We investigate methods that analyze the relationships of users as well as how their posts influence others in the network. We aim to devise scores that quantify topic based influence of each user. The effectiveness of the methods will be tested using data from micro-blogging service Twitter.


DATE: 16 March, 2015, Monday @ 16:15