Event Causality Extraction from Natural Science Literature

We aim to develop a text mining framework capable ofidentifying and extractingcausal dependenciesamongchanging variables(orevents) from scientific publications in the cross-disciplinary field ofoceanographic climate science. The extracted information can be usedto infer new knowledge or to find out unknown hypotheses throughreasoning, which forms the basis of a knowledge discovery supportsystem. Automatic extraction of causal knowledge from text contentis a challenging task. Generally, the approaches of causal relationidentification proposed in the literature target specific domain such asonline news or biomedicine as the domain has significant influence oncausality expressions found in the domain texts. Therefore, the existingmodels of causality extraction may not be directly portable to other/newdomains. In this paper, we describe the nature of causation observed inclimate science domain, review the state-of-the-art approaches in causalknowledge extraction from text and carefully select the methods andresources most likely to be applicable to the considered domain