Measuring Oral Reading Fluency at Scale

Speaker:
Organiser:
Abhishek Sinha
Date:
Friday, 11 Dec 2026, 14:00 to 14:45
Venue:
Homi Bhabha Auditorium
Abstract

Foundational literacy and numeracy are the cornerstones of India’s New Education Policy (NEP 2020) which also recommends competency-based learning where assessments play a critical role. A key indicator of reading ability is oral reading fluency (ORF), traditionally used to evaluate critical literacy proficiencies including language understanding. Given the time and resource intensive nature of such assessment, it is of interest to automate the measurement of reading fluency.  We discuss the considerations that arise in view of speech processing paradigms that range from knowledge-based feature engineering to deep learning from labeled data. In the case of automated educational assessments, both, prediction performance and the interpretability of outcomes, are critical. We present a journey through field data collection, computational modeling of pedagogical rubrics and working around limitations of state-of-the-art algorithms in the real world to achieve an objective and reliable implementation of ORF measurement at scale. The TARA system, developed at IIT Bombay, was endorsed last year by the Ministry of Education for country-wide deployment of foundational literacy assessment (https://www.youtube.com/watch?v=VrfrgS9JK_Q)

Bio: Preeti Rao was on the faculty of Electrical Engineering at I.I.T. Bombay until June 2026, teaching and researching in the area of signal processing with applications in speech and audio.  She was a recipient of the Abdul Kalam Technology Innovation National Fellowship (2020-2025) for work on the development of instrumental measures for oral skill.  Post retirement now, she leads a voice AI edutech start-up, looking forward to contributing her mite to the literacy and language learning ecosystem in the country.