Ved has been reading for his DPhil at the Department of Engineering Science and Lady Margaret Hall since 2022. His research is the field of Natural Language Processing and more specifically on the extraction of events and their relationships from text. He is supervised by Janet Pierrehumbert.
He is interested in the epistemic power of modals, adverbs, adjectives and speech-act verbs. In the example, ‘John thinks Mary may be at her office,’ the verb ‘thinks’ and the modal ‘may’ conveys to the listener that they should take it that there is a high chance Mary is not at her office. The research involves exploring the ability of modern language models and neuro-symbolic methods that use traditional linguistic features along with language models to estimate the confidence of propositions and events modified by these articles and present in the sub-clauses of speech act verbs.
Ved is also currently employed on a research project on Narrative Detection and Tracking using NLP at the Economics Department of the University of Oxford. The project is funded by the Alan Turing Institute and the Defence Science and Technology Laboratory. As part of the project he is exploring the graphical modelling of narratives.
Before starting his DPhil, Ved completed his Bachelors of Technology at the Manipal Institute of Technology in Mechatronics at Manipal and his Masters of Technology at the International Institute of Information Technology – Bangalore. He has also worked in multiple startups for 5 years. The longest period at Slang Labs, where he was a Senior NLP Engineering working on the NLP engine powering in-app voice assistants.