Placing (Historical) Events on a Timeline: A Classification cum Co-ref Resolution Approach

Anonymous

Abstract: The event timeline provides one of the most effective ways to visualize the important historical events that occurred over a period of time, presenting the insights that may not be so apparent from reading the equivalent information in textual form. By leveraging generative adversarial learning for important event classification and by assimilating knowledge based tags for improving the performance of event coreference resolution we introduce a two staged system for event timeline generation from multiple (historical) text documents. In addition, we propose a vis-timeline based visualization technique to portray the event timeline. We demonstrate our results on two very well known historical documents -- the Collected Works of Mahatma Gandhi (CWMG) and the Collected Works of Abraham Lincoln (CWAL). Our results can be extremely helpful for historians, in advancing research in history and in understanding the socio-political landscape of a country as reflected in the writings of political leaders/scholars. Our work has some parallels with timeline summarization (TLS) tasks and therefore we use these as baselines. Rigorous experiments demonstrate that prior event detection which was hitherto absent in the TLS methods} can improve summarization performance. In order to show that our methods are very generic we reuse our method to visualize the evolution of coronavirus related events in India from a collection of various COVID-19 articles.