Borna Jafarpour

Borna Jafarpour

Research Scientist

FULL NAME + PROFESSIONAL DESIGNATION
Borna Jafarpour, PhD in Computer Science

ROLE
Research Scientist

TR LOCATION
Toronto

FOCUS AREA
Natural Language Processing

How did you get into this area of work? (Was there a specific moment where you said this is what I want to do?)
Online abuse and bullying are negative and unintended consequences of internet popularity. I was involved in research to improve how online abuse can be detected in order to protect the users. We were surprised by how well it worked even for cases that we had not anticipated. That project showed me how vibrant the Natural Language Processing (NLP) research domain is and how much of an impact it can have on our daily lives.

What kind of projects do you work on?
We work on cutting edge Deep Learning as well as classic Artificial Intelligence (AI) techniques such as Language Modelling, Active Learning and Transfer Learning to solve problems in the legal domain which will help legal professionals do their research more efficiently. This will eventually bring legal services’ costs down and will make such services more affordable to the public. Legal domain is a niche research area which makes it a very challenging and rewarding area to work in.

What is one interesting fact you’ve come across while performing your work?This is my first experience in the legal domain, and I was surprised to see how much AI has influenced the legal business and the extent to which Thomson Reuters has led this effort.

Last thing you attended/ate/drank – and how would you apply AI/ML to improve that experience?
Generative models are becoming more powerful in generating interesting pictures and music, for example. I attended a standup comedy event recently and that gave me the idea that generative models can be used for creative writing. I believe as language generation techniques get more sophisticated, they’ll be able to help writers, standup comedians, poets and even politicians to do their jobs more efficiently.

Any AI/ML related predictions for the next couple of years?
I expect natural language modelling will get so powerful that will be able to create jokes.

Posts by this author