Extraction of causality and related events using text analysis

Identification and extraction of causality from the text have always been an active domain of research since long. Along with the identified causalities, it is also important to identify the associated causal events. Modelling of a system that can automatically extract the causalities becomes more tedious with the need of a large annotated corpus to train the model. Often the performance of the system is influenced by the application domain and the terminologies that are vital and specific to the domain itself. Through this paper, we propose a process framework that learns to extract causality and other relevant information from the text input through applied Machine learning. The test data that we used is from the domain of social science where the text contains the narratives from women about the experiences of their life. The results are represented as a causal graph model which can be further used for visualization and interpretation of causality.