Adaptive Topic Follow-Up on Twitter

Abstract

Twitter provides a strictly limited API that makes it difficult for a simple search using pre-defined textual patterns to provide satisfying coverage of the topic of interest. This paper discusses a tweet acquisition system, that queries Twitter API using a set of key phrases, then analyzes the retrieved tweets. In order to achieve better coverage of the searched topic, the system employs an adaptive query generation mechanism that iteratively enriches the set of textual relevant patterns based on the previously collected tweets using an explore-exploit strategy. The paper also demonstrates an application called Topic Follow-up on Twitter (TFT) that is built on top of the acquisition system and aims at linking tweets with online articles. It first extracts a set of key phrases from the submitted news article and then utilizes the acquisition and analysis components of the system. Using this application, we will show how the adaptive searching mechanism of the tweet acquisition system improves the coverage of the topic of interest. Video: http://bit.ly/2kqkikB.

DOI: 10.1109/ICDE.2017.188

2 Figures and Tables

Cite this paper

@article{Alsaudi2017AdaptiveTF, title={Adaptive Topic Follow-Up on Twitter}, author={Abdulrahman Alsaudi and Mehdi Sadri and Yasser Altowim and Sharad Mehrotra}, journal={2017 IEEE 33rd International Conference on Data Engineering (ICDE)}, year={2017}, pages={1385-1386} }