Varun Shijo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Varun Shijo has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Anarghya Das
University at Buffalo, Buffalo, New York, USA
1 shared publication - Chuhui Liu
Department of Computer Science and Engineering, University at Buffalo, Buffalo, New York, USA
1 shared publication - Ling-Yu Guo
Department of Communicative Disorders and Sciences, University at Buffalo, NY
1 shared publication - Long Cao
Department of Computer Science and Engineering, University at Buffalo, Buffalo, New York, USA
1 shared publication - Shuwei Hou
Department of Computer Science and Engineering, University at Buffalo, Buffalo, New York, USA
1 shared publication - Wei Bo
University at Buffalo, Buffalo, New York, USA
1 shared publication
Latest Publications
- Open Access
- 30 September 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Vol. 10 (3) , 1-37
Speech Annotation and Transcription Enhancer (SATE): An Automated System for Child Language Sample Analysis
Developmental language disorder affects approximately 7-10% of children in the United States and is associated with long-term difficulties in literacy, academic achievement, and social participation. Language sample analysis (LSA) is widely regarded as a gold-standard approach for evaluating children's language development; yet, it remains underused in clinical and research practice because it depends on labor-intensive and time-consuming manual transcription and coding. Therefore, we propose SATE, a scalable and automated system for child language assessment. It implements a modular machine learning pipeline that reproduces the full LSA workflow, including verbatim transcription, utterance segmentation, speaker identification, maze annotation, lexical and grammatical analyses, and language metrics computation. The pipeline is complemented by a user interface that supports efficient review and correction of model output, allowing researchers and clinicians to validate and refine automated results. Experimental evaluation on child speech corpora demonstrates that SATE substantially outperforms existing baselines, achieving 11.81% word error rate and strong correlations with authenticated results of language metrics. A counterbalanced human-in-the-loop study showed that SATE-assisted LSA reduced task completion time by 55.3% while maintaining comparable accuracy to manual LSA, and usability questionnaires yielded a mean SUS score of 81.25. These results provide evidence that SATE is a practical tool for child language sample analysis.
- 30 September 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Vol. 10 (3) , 1-37
Open AccessSpeech Annotation and Transcription Enhancer (SATE): An Automated System for Child Language Sample Analysis
Developmental language disorder affects approximately 7-10% of children in the United States and is associated with long-term difficulties in literacy, academic achievement, and social participation. Language sample analysis (LSA) is widely regarded as a gold-standard approach for evaluating children's language development; yet, it remains underused in clinical and research practice because it depends on labor-intensive and time-consuming manual transcription and coding. Therefore, we propose SATE, a scalable and automated system for child language assessment. It implements a modular machine learning pipeline that reproduces the full LSA workflow, including verbatim transcription, utterance segmentation, speaker identification, maze annotation, lexical and grammatical analyses, and language metrics computation. The pipeline is complemented by a user interface that supports efficient review and correction of model output, allowing researchers and clinicians to validate and refine automated results. Experimental evaluation on child speech corpora demonstrates that SATE substantially outperforms existing baselines, achieving 11.81% word error rate and strong correlations with authenticated results of language metrics. A counterbalanced human-in-the-loop study showed that SATE-assisted LSA reduced task completion time by 55.3% while maintaining comparable accuracy to manual LSA, and usability questionnaires yielded a mean SUS score of 81.25. These results provide evidence that SATE is a practical tool for child language sample analysis.