From Audio to Semantics: Approaches to End-to-End Spoken Language Understanding
| Autor: | Arun Narayanan, Galen Chuang, Zhongdi Qu, Rohit Prabhavalkar, Neeraj Gaur, Parisa Haghani, Pedro J. Moreno, Michiel Bacchiani, Austin Waters |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
FOS: Computer and information sciences
Sound (cs.SD) Computer science Natural language understanding Word error rate 010501 environmental sciences computer.software_genre Semantics 01 natural sciences Computer Science - Sound Domain (software engineering) Set (abstract data type) 030507 speech-language pathology & audiology 03 medical and health sciences Audio and Speech Processing (eess.AS) Argument FOS: Electrical engineering electronic engineering information engineering 0105 earth and related environmental sciences Computer Science - Computation and Language business.industry Task analysis Artificial intelligence 0305 other medical science business Computation and Language (cs.CL) computer Natural language processing Electrical Engineering and Systems Science - Audio and Speech Processing Spoken language |
| Zdroj: | SLT |
| DOI: | 10.1109/slt.2018.8639043 |
| Popis: | Conventional spoken language understanding systems consist of two main components: an automatic speech recognition module that converts audio to a transcript, and a natural language understanding module that transforms the resulting text (or top N hypotheses) into a set of domains, intents, and arguments. These modules are typically optimized independently. In this paper, we formulate audio to semantic understanding as a sequence-to-sequence problem [1]. We propose and compare various encoder-decoder based approaches that optimize both modules jointly, in an end-to-end manner. Evaluations on a real-world task show that 1) having an intermediate text representation is crucial for the quality of the predicted semantics, especially the intent arguments and 2) jointly optimizing the full system improves overall accuracy of prediction. Compared to independently trained models, our best jointly trained model achieves similar domain and intent prediction F1 scores, but improves argument word error rate by 18% relative. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|