Summarization of Historical Articles Using Temporal Event Clustering

James Gung1 and Jugal Kalita2
1Miami University, 2University of Colorado, Colorado Springs


Abstract

In this paper, we investigate the use of temporal information for improving extractive summarization of historical articles. Our method clusters sentences based on their timestamps and temporal similarity. Each resulting cluster is assigned an importance score which can then be used as a weight in traditional sentence ranking techniques. Temporal importance weighting offers consistent improvements over baseline systems.