Industry Scale Semi-Supervised Learning for Natural Language Understanding

Abstract

This paper presents a production Semi-Supervised Learning (SSL) pipeline based on the student-teacher framework, which leverages millions of unlabeled examples to improve Natural Language Understanding (NLU) tasks. We investigate two questions related to the use of unlabeled data in the production SSL context: 1) how to select samples from a huge unlabeled data pool that are beneficial for SSL training, and 2) how do the selected data affect the performance of different state-of-the-art SSL techniques. We compare four widely used SSL techniques, Pseudo-Label (PL), Knowledge Distillation (KD), Virtual Adversarial Training (VAT), and Cross-View Training (CVT) in conjunction with two data selection methods including committee-based selection and submodular optimization-based selection. We further examine the benefits and drawbacks of these techniques when applied to intent classification (IC) and named entity recognition (NER) tasks in the English language using a public dataset (SNIPS) and real-world data from Amazon Alexa. To conclude we provide guidelines specifying when each of these methods might be beneficial to improve large-scale NLU systems.