Shekhar has a Ph.D. in computer science as well as in philosophy, and teaching experience in a range of topics and institutions. We are delighted to have him join us in DSI. We asked him these questions:
What types of data do you typically work with in your research? How has the scale of data changed over the course of your career?
I typically work with textual data. My work has been in creating models for natural language understanding. Of course, each year more text is being generated at dizzying speeds. But that is really not very interesting. What is interesting is the way massive amounts of textual data are being used to create what are called “word embeddings” (vector representation of words) and how these vector embeddings are used as inputs for creating deep learning models that have produced very impressive breakthroughs in natural language understanding.
What methods do you use when working with data?
I use mostly deep learning models such as transformers when working with linguistic data.
What social challenges motivate your work?
They say information is power and having access to more data empowers citizens. But the truth is that it only empowers those who have the technical and/or financial resources to mine this data and draw insights from them. So, paradoxically, making more data or information available to people has the potential to deepen the inequalities that exist in society. My work is motivated by the desire to reduce these inequalities by building tools that provide anyone the ability to acquire the information and insights they need without having to spend much money or to acquire technical expertise.
What are your hopes for data science at Brown?
My hope is that more and more constituencies in Brown begin to use data in their work and their courses and utilize the services of DSI to help them do that.
What’s your favorite thing about life in Rhode Island?
Ask me that in a year or so.