Robust Spoken Language Understanding via Paraphrasing
This addresses a critical issue for SLU and dialog systems by enhancing their ability to handle rare and complex paraphrased inputs, though it is incremental as it builds on existing SLU models.
The paper tackles the problem of performance degradation in spoken language understanding (SLU) systems when encountering paraphrased utterances and out-of-vocabulary words, by introducing a novel paraphrasing-based SLU model that improves robustness, as demonstrated in experiments on benchmark and in-house datasets.
Learning intents and slot labels from user utterances is a fundamental step in all spoken language understanding (SLU) and dialog systems. State-of-the-art neural network based methods, after deployment, often suffer from performance degradation on encountering paraphrased utterances, and out-of-vocabulary words, rarely observed in their training set. We address this challenging problem by introducing a novel paraphrasing based SLU model which can be integrated with any existing SLU model in order to improve their overall performance. We propose two new paraphrase generators using RNN and sequence-to-sequence based neural networks, which are suitable for our application. Our experiments on existing benchmark and in house datasets demonstrate the robustness of our models to rare and complex paraphrased utterances, even under adversarial test distributions.