Edinburgh Research Explorer

WERd: Using Social Text Spelling Variants for Evaluating Dialectal Speech Recognition

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Original languageEnglish
Title of host publicationIEEE Automatic Speech Recognition and Understanding Workshop (ASRU 2017)
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Number of pages8
ISBN (Electronic)978-1-5090-4788-8
ISBN (Print)978-1-5090-4789-5
Publication statusPublished - 25 Jan 2018
Event2017 IEEE Automatic Speech Recognition and Understanding Workshop - Okinawa, Japan
Duration: 16 Dec 201720 Dec 2017


Conference2017 IEEE Automatic Speech Recognition and Understanding Workshop
Abbreviated titleASRU 2017
Internet address


We study the problem of evaluating automatic speech recognition (ASR) systems that target dialectal speech input. A major challenge in this case is that the orthography of dialects is typically not standardized. From an ASR evaluation perspective, this means that there is no clear gold standard for the expected output, and several possible outputs could be considered correct according to different human annotators, which makes standard word error rate (WER) inadequate as an evaluation metric. Such a situation is typical for machine translation (MT), and thus we borrow ideas from an MT evaluation metric, namely TERp, an extension of translation error rate which is closely-related to WER. In particular, in the process of comparing a hypothesis to a reference, we make use of spelling variants for words and phrases, which we mine from Twitter in an unsupervised fashion. Our experiments with evaluating ASR output for Egyptian Arabic, and further manual analysis, show that the resulting WERd (i.e., \emph{WER for dialects}) metric, a variant of TERp, is more adequate than WER for evaluating dialectal ASR.


2017 IEEE Automatic Speech Recognition and Understanding Workshop


Okinawa, Japan

Event: Conference

Download statistics

No data available

ID: 43692166