Fault Knowledge Acquisition of Aircraft Based on Event Extraction Technology.

In order to make full use of aircraft fault information of total life cycle and construct a knowledge base for intelligent fault diagnosis, this paper preliminarily explored extracting fault knowledge from massive unstructured texts by using the event extraction technology. According to the business requirements of fault diagnosis, we defined the elements of fault events, including trigger words, and fault time, argument roles such fault time, fault occasion, fault unit, signal indicate. To extract above elements, three models, including fault event identification model, trigger extraction model and Argument extraction model is developed in this paper. The results show that this method is effective.