Research in computer systems and platforms, corpus linguistics, computational linguistics, speech processing and machine learning

We develop cutting-edge approaches and technologies for automated assessment that enhance the experience of learning English language and benefit learners and teachers worldwide.

Founded in 2013, the Cambridge Institute for Automated Language Teaching and Assessment (ALTA) brings together language learning and assessment experts from Cambridge University Press & Assessment and AI-specialists from the departments of Computer Science and Technology and Engineering of the University of Cambridge, organised into a consortium of projects.

Our primary objective is to develop cutting-edge AI technologies that enhance the experience of English language teaching, learning, and assessment and benefit teachers and learners worldwide. We conduct research into improving linguistic proficiency of learners and automated forms of assessment of language performance. We do this by using techniques from Machine Learning and Natural Language Processing.

We work with principal investigators from the Departments of Engineering and Theoretical and Applied Linguistics. We are part of the Cambridge University's Language Sciences Interdisciplinary Research Centre.

Sponsorship from Cambridge University Press & Assessment supports a team of PhD students and post-doctoral researchers who investigate a range of topics, including:

• Language Assessment
• Teaching Feedback
• Adaptive Learning
• Content Creation and Calibration
  

To advance research in these fields, we must delve deeper into the following technologies:

• Text and speech processing
• Machine learning and natural language processing
• Fine-tuning Large Language Models (LLMs)
• Security, platforms and deployment

We draw from Cambridge University Press & Assessment’s extensive experience in using technology to support language learning and assessment, including the multi-million-word Cambridge Learner Corpus. This is a uniquely large database of learners' English which is used to enhance Cambridge English exams and learning materials. Huge advances in areas like speech recognition and machine learning mean that automated tools can now complement the work of human assessors, giving accurate and reliable evaluations of language, as well as providing actionable feedback to teachers and learners on areas for improvement. Alongside the University of Cambridge, ALTA and the other departments, we are together carrying out a lot of ground-breaking work adding a new dimension to language teaching and learning.

ALTA research is supported by ELiT (English Language iTutoring) through data gathering and the deployment of AI technologies into Cambridge University Press & Assessment’s learning and assessment products.

Language acquisition is a complex and distinctively human process. We believe that learners get better outcomes when they are supported by teachers. Similarly, we believe that language assessment is more effective when informed by the expertise provided by our trained examiners.

AI technology, including the use of automated tools, powered by machine learning, can bring huge opportunities to support teachers and learners. However, we believe in a responsible and ethical use in the educational context with rigorous evaluations and careful human invigilation as a standard of good practice.