Spoken SQuAD: A Study of Mitigating the Impact of Speech Recognition Errors on Listening Comprehension
This addresses the challenge of understanding spoken content for applications like multimedia access, though it is incremental as it adapts an existing text-based dataset to speech.
The paper tackles the problem of machine listening comprehension by introducing Spoken SQuAD, a new task based on speech, and finds that speech recognition errors severely degrade performance, proposing methods to mitigate this impact.
Reading comprehension has been widely studied. One of the most representative reading comprehension tasks is Stanford Question Answering Dataset (SQuAD), on which machine is already comparable with human. On the other hand, accessing large collections of multimedia or spoken content is much more difficult and time-consuming than plain text content for humans. It's therefore highly attractive to develop machines which can automatically understand spoken content. In this paper, we propose a new listening comprehension task - Spoken SQuAD. On the new task, we found that speech recognition errors have catastrophic impact on machine comprehension, and several approaches are proposed to mitigate the impact.