Dr Alexandra Birch
In the following table, contact information relevant to the page. The first column is for visual reference only. Data is in the right column.
|Job title:||Innovation Fellow|
|Division:||School of Informatics|
|Organisation:||University of Edinburgh|
|Tags:||Early Career Forum, Fellowship: Innovation Fellowship, Researcher, University of Edinburgh|
I work in the field of machine translation and recently I have focussed on neural machine translation where advances using sub-word units and monolingual data have beaten state-of-the-art baselines. More generally I am interested in how computational linguistics and machine learning can be combined to provide compelling NLP applications
Neural machine translation (NMT) has recently made major advances in translation quality and this technology has been rapidly adopted by industry leaders. However, high performing neural models require millions of human translated sentences for training. For many real-world applications, there is not enough data to build useful MT systems. In this project I will stretch the resources and capabilities that we have, in order to develop robust MT technologies which are capable of being deployed for low-resource language pairs and for highly specialised low-resource domains.
I will investigate making translation significantly more robust by using the intuition that translated (or parallel) corpora contain enormous redundancies, and are an inefficient way to learn to translate. Inspired by human learning, we will study models which build up meaning compositionally and are able to learn to learn, thus creating models which only need a few training examples. We will also develop machine learning techniques, such as transfer learning and data augmentation, to extract knowledge from monolingual and parallel resources from other languages and domains.
My team will provide translations for language pairs which were not previously well served by automatic machine translation. This will allow our partners, BBC World Service, to cover under-resourced languages, helping them to deliver impartial reporting across the world. In the long term, this project will have a wider impact on British industry by breaking down language barriers affecting international trade, and by significantly improving the quality and resilience of transformative AI language technologies.
The fellowship allows me to establish myself as a research and innovation leader in the language technology space. I have a wealth of experience at leading large projects with industry. I am ready to embark on more focused knowledge transfer with key players that (1) I already work with and (2) where my work will have a high commercial impact.
The fellowship allows me to:
1) Further develop my research group, and extend the Edinburgh machine translation group, in order to sustain and enhance our world-leading research.
2) Successfully transfer robust, state-of-the-art technology into industry and drive uptake of artificial intelligence in the BBC and the media industry.
3) Further develop my network of national and international collaborations, create joint research outputs and initiate new international research projects.