Event recognition as an important part of the event study, get more and more attention. To handle the problems of single feature selection and the weakling to a recognition model because of irrelevant and redundant features in a event recognition, an event recognition method uniting multi-features was proposed. This method first relies on LTP for processing the lexical features, syntactic features, semantic role features and semantic dependency and other characteristics of the raw data, and then use Relief algorithm to select features to construct the feature vector, and finally use the SVM classification to classify the feature vector to event recognition. The experimental results demonstrate that the recognition methods have achieved good results.
CITATION STYLE
Liu, T., Chao, Y., Zhang, S., & Liao, T. (2020). Research on Event Recognition Based on Multi-feature Fusion. In Advances in Intelligent Systems and Computing (Vol. 1017, pp. 1189–1197). Springer Verlag. https://doi.org/10.1007/978-3-030-25128-4_150
Mendeley helps you to discover research relevant for your work.