back to list

Project: Mining Causal Relations Between Real-world Human Events

Description

Motivation: 

The ACLED dataset (https://acleddata.com/knowledge-base/codebook/) provides a detailed record of political violence and protest events, capturing actors, timelines, and descriptions of the incidents. However, this rich data remains largely underutilized when it comes to understanding the causal relationships between events. 

While common knowledge graphs typically rely on actor-based connections to link events (e.g., Event A and Event B are related because the same actor is involved), they fall short when it comes to analyzing why one event happened in the context of the preceding or triggering events. This gap limits our understanding of the complex dynamics that drive political violence and social unrest. 

Model and identify causal relations between events would significantly enhance our understanding of these dynamics. Such models could reveal how certain events lead to others, allowing us to construct causal chains or graphs of events. For example, rather than simply observing that two protest movements involve the same actor, we could discern whether one protest was a direct consequence of a previous one (e.g., due to failure of negotiations, repression, etc.).

Impact:

This project can provide a deeper understanding of conflict dynamics, help predictive modeling and forecasting, scientific contributions to textual causality study and NLP.

Methodology:

This project consists of several key steps, including 

1) Causal Relation Detection among text data.

2) Matching causes to historical events. Consider temporal and geographical constrains, and the way to deal with vague or multiple links.

3) Evaluation design. Since no ground truth is provided, human evaluation or LLM based evaluation?

Different possible research objectives can be discussed and chosen, dependent on the interest of the student. The started date is flexible but earlier is better. For a successful thesis, there will be the possibility to work with the supervisor on a publication.

If you are interested, please contact:

* Supervisor: Dr. Deng

* Email: s.deng@tue.nl 

* Office: MF 7.145

Details

Student

TL

Tianyuan Liu

Supervisor

Amy Deng