Automatic scoring of pronunciation quality
Speech Communication
Automatic Pronunciation Scoring for Language Instruction
ICASSP '97 Proceedings of the 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP '97)-Volume 2 - Volume 2
Automatic scoring of non-native spontaneous speech in tests of spoken English
Speech Communication
HLT '11 Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies - Volume 1
Detecting structural events for assessing non-native speech
IUNLPBEA '11 Proceedings of the 6th Workshop on Innovative Use of NLP for Building Educational Applications
Exploring content features for automated speech scoring
NAACL HLT '12 Proceedings of the 2012 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Hi-index | 0.00 |
This study presents a method that assesses ESL learners' vocabulary usage to improve an automated scoring system of spontaneous speech responses by non-native English speakers. Focusing on vocabulary sophistication, we estimate the difficulty of each word in the vocabulary based on its frequency in a reference corpus and assess the mean difficulty level of the vocabulary usage across the responses (vocabulary profile). Three different classes of features were generated based on the words in a spoken response: coverage-related, average word rank and the average word frequency and the extent to which they influence human-assigned language proficiency scores was studied. Among these three types of features, the average word frequency showed the most predictive power. We then explored the impact of vocabulary profile features in an automated speech scoring context, with particular focus on the impact of two factors: genre of reference corpora and the characteristics of item-types. The contribution of the current study lies in the use of vocabulary profile as a measure of lexical sophistication for spoken language assessment, an aspect heretofore unexplored in the context of automated speech scoring.