A spoken language understanding approach using successive learners

In this paper, we describe a novel spoken language understanding approach using two successive learners. The first learner is used to identify the topic of an input utterance. With the restriction of the recognized target topic, the second learner is trained to extract the corresponding slot-value pairs. The advantage of the proposed approach is that it is mainly datadriven and requires only minimally annotated corpus for training whilst retaining the understanding robustness and deepness for spoken language. Experiments have been conducted in the context of Chinese public transportation information inquiry domain. The good performance demonstrates the viability of the proposed approach.