Ensemble learning for question classification

In this paper, a new method for question classification is proposed, which employs ensemble learning algorithms to train multiple question classifiers. These component learners are combined to produce the final hypothesis. In detail, the feature spaces are obtained through extracting high-frequency keywords from questions corpus and the method of word semantic similarity is performed to adjust the feature weights. The ensemble methods, Bagging and AdaBoost, are applied to construct an ensemble of decision trees to tackle the problem of question classification respectively. Experiments on the Chinese question system of tourism domain show that the ensemble methods could effectively improve the classification accuracy.