Skip to main content

IEOR Seminar by Diptesh Kanojia

Title: Natural Language Processing

Date, Time: Monday, 21st October 2019, 4.00 PM - 5.30 PM

Venue: IEOR Teaching Lab (IE 011), Ground floor, IEOR Building.

Speaker: Diptesh Kanojia (Research Scholar, IITB - Monash Research Academy)

Abstract:  In the talk, we discuss how NLP is at the crux of many AI problems that need to be solved. We start with a brief introduction of the different NLP sub-tasks and go ahead to discuss in detail a few of which we have worked on. We discuss the generality of the Machine Translation task along with some specific tasks such as Cognate detection, Sarcasm Detection, and Computational Phylogenetics. Mostly we describe the problems associated with these tasks and later in discussions, we can ponder upon the possible solutions which can be helpful in resolving the current issues. The main purpose of this talk is to briefly describe to the audience what NLP tasks are 'out-there' and helps them 'zero-in' on one or more which they would like to work with. The talk also discusses the tangential area of cognitive psycholinguists and how it is helping tackle the daunting problems in various NLP sub-tasks like Sarcasm Detection and Sentiment Analysis.

About the Speaker: Diptesh Kanojia is a Ph.D. research scholar at the IITB-Monash Research Academy also affiliated to the Center for Indian Language Technology (CFILT) Lab at IIT Bombay. He has about nine years of experience in NLP research under the guidance of Prof. Pushpak Bhattacharyya (CSE, IITB/IITP). He has worked on various facets of NLP research including but not limited to Cognate Detection, Computational Phylogenetics, Sentiment Analysis, Cognitive Psycholinguistics, and Machine Translation. His current research on Cognate Detection and Phylogenetics is being advised by Prof. Pushpak, Prof. Malhar Kulkarni (HSS, IITB), and Prof. Reza Haffari (Monash University). He is a graduate of Uttar Pradesh Technical University (B.Tech.) and was a Research Staff at the CFILT Lab for a period of 3 years.

 

News Category
Date Posted