Modeling Confidence in Sequence-to-Sequence Models
This addresses the need for reliable confidence estimation in sequence-to-sequence models for applications like machine translation and speech recognition, though it is incremental as it builds on existing alignment and probability methods.
The paper tackles the problem of assessing output quality in neural sequence-to-sequence models by using similarity between training and test conditions as a confidence measure, showing improvements in machine translation and achieving 60% error detection by reviewing 20% of data in automatic speech recognition.
Recently, significant improvements have been achieved in various natural language processing tasks using neural sequence-to-sequence models. While aiming for the best generation quality is important, ultimately it is also necessary to develop models that can assess the quality of their output. In this work, we propose to use the similarity between training and test conditions as a measure for models' confidence. We investigate methods solely using the similarity as well as methods combining it with the posterior probability. While traditionally only target tokens are annotated with confidence measures, we also investigate methods to annotate source tokens with confidence. By learning an internal alignment model, we can significantly improve confidence projection over using state-of-the-art external alignment tools. We evaluate the proposed methods on downstream confidence estimation for machine translation (MT). We show improvements on segment-level confidence estimation as well as on confidence estimation for source tokens. In addition, we show that the same methods can also be applied to other tasks using sequence-to-sequence models. On the automatic speech recognition (ASR) task, we are able to find 60% of the errors by looking at 20% of the data.