Joint Learning of Domain Classification and Out-of-Domain Detection with Dynamic Class Weighting for Satisficing False Acceptance Rates
This work addresses the need for more efficient and accurate domain classification in spoken dialog systems, reducing user confusion and interaction costs, though it appears incremental by building on existing methods.
The paper tackles the problem of domain classification and out-of-domain detection in spoken dialog systems by introducing a neural joint learning model with dynamic class weighting, which significantly improves domain classification accuracy while meeting target false acceptance rates.
In domain classification for spoken dialog systems, correct detection of out-of-domain (OOD) utterances is crucial because it reduces confusion and unnecessary interaction costs between users and the systems. Previous work usually utilizes OOD detectors that are trained separately from in-domain (IND) classifiers, and confidence thresholding for OOD detection given target evaluation scores. In this paper, we introduce a neural joint learning model for domain classification and OOD detection, where dynamic class weighting is used during the model training to satisfice a given OOD false acceptance rate (FAR) while maximizing the domain classification accuracy. Evaluating on two domain classification tasks for the utterances from a large spoken dialogue system, we show that our approach significantly improves the domain classification performance with satisficing given target FARs.