摘要

This paper investigates the problem on how to track a microblog user's trail effectively and accurately by using the content of the user's posts. A framework for modeling and extracting location information and time-stamp from Sina Weibo is presented. In this framework, we propose and evaluate a probabilistic model for estimating a user's location in an energy-efficient way. Relying on this framework, current available visualization techniques can be easily applied to show the person tacking information. Finally, a system based on the proposed model is developed and verified. The experiment results show that this system can automatically and accurately mine a person's trail based purely on the content of the user's tweets.

  • 出版日期2015

全文