Efficient Keyword-Related Data collection in a Social Network with Weighted Seed Selection

摘要

Data mining in a social network can yield interesting perspectives to understanding human behavior or detecting topics or communities. However, it is difficult to gather the data related to a specific topic due to the main characteristics of social media data: large, noisy, and dynamic. To collect the data related to a specific topic efficiently, we propose a new algorithm that selects better seeds with limited resources. Furthermore, we compare two data sets collected by the algorithm and existing approaches.