Automated Scoring of Nonnative Speech Using the SpeechRaterSM v. 5.0 Engine. Issue 1 (12th April 2018)

• Record Type: Journal Article
• Authors: Chen, Lei; Zechner, Klaus; Yoon, Su‐Youn; Evanini, Keelan; Wang, Xinhao; Loukina, Anastassia; Tao, Jidong; Davis, Lawrence; Lee, Chong Min; Ma, Min; Mundkowsky, Robert; Lu, Chi; Leong, Chee Wee; Gyawali, Binod
• Is Part Of: ETS research report series. Issue 1(2018)
• Page Start: 1
• Page End: 31
• Publication Date: 2018-04-12
• Subjects: Automated speech scoring -- automated speech recognition -- scoring models -- natural language processing -- English assessments; Universities and colleges -- Entrance examinations; Universities and colleges -- Graduate work -- Examinations; Universities and colleges -- United States -- Entrance examinations; Universities and colleges -- United States -- Graduate work -- Examinations; Graduate Record Examination; Educational tests and measurements; Social sciences; Education -- Research; Education -- Research; Educational tests and measurements; Graduate Record Examination; Social sciences; Universities and colleges -- Entrance examinations; Universities and colleges -- Graduate work -- Examinations; United States; 378
• DOI: 10.1002/ets2.12198
• Languages: English
• ISSNs: 2330-8516

Abstract

Abstract: This research report provides an overview of the R&D efforts at Educational Testing Service related to its capability for automated scoring of nonnative spontaneous speech with the SpeechRater SM automated scoring service since its initial version was deployed in 2006. While most aspects of this R&D work have been published in various venues in recent years, no comprehensive account of the current state of SpeechRater has been provided since the initial publications following its first operational use in 2006. After a brief review of recent related work by other institutions, we summarize the main features and feature classes that have been developed and introduced into SpeechRater in the past 10 years, including features measuring aspects of pronunciation, prosody, vocabulary, grammar, content, and discourse. Furthermore, new types of filtering models for flagging nonscorable spoken responses are described, as is our new hybrid way of building linear regression scoring models with improved feature selection. Finally, empirical results for SpeechRater 5.0 (operationally deployed in 2016) are provided. Abstract : Report Number: ETS RR–18‐10