Jasneet Sabharwal

Chief Innovation Officer @ Text IQ

jasneet _at_ textiq.com

Feb 2021 - Present
Text IQ: Chief Innovation Officer
Feb 2017 - Present
Text IQ: Director of Engineering
May 2016 - Jan 2017
Text IQ: R&D Engineer
May 2016 - Present
Text IQ Labs Canada: Director
Fall 2015
Text IQ: Research Engineer Intern
March 2015
CODE 2015 Winner
Best Student Team award at Canadian Open Data Experience 2015 for building a non-linear visualization tool for finding a post-secondary education institute in Canada.
November 2014 - March 2015
WorkSafeBC: Mitacs-Accelerate Graduate Research Intern
Detecting Trends in Conversations using Natural Language Processing and Visualization.
Summer 2014
Maluuba Inc: NLP Development Intern
Supervised classification and information extraction from natural language queries. Emphasised on feature engineering, feature reduction, model training & fine tuning and post-processing.
September 2013 - May 2016
Natural Language Lab, Simon Fraser University: Research Assistant
Visualization of historical human events using dimensionality reduction techniques.
September 2013 - May 2016
Simon Fraser University: MSc. Computing Science
Specializing in Natural Language Processing (emphasis on Machine Learning application). My advisors are Anoop Sarkar and Fred Popowich.
May 2012 - July 2013
Germin8: Software Engineer (Analysis Engine Research Team)
Sentiment Analsysis, Named Entity Recognition, Intention Analysis, Realtime Distributed Analysis Framework.
March 2011 - May 2012
NextGen Invent Corporation: Software Engineer (PureAnalyzer Team)
Sentiment Analsysis, Web Crawling, Big Data Indexing and Searching.
2006 - 2010
Guru Gobind Singh Indraprastha University: Bachelor of Technology
Majored in Computer Science and Engineering

Publication

End-to-End Sentiment Analysis of Twitter Data
We presented an end-to-end pipeline for sentiment analysis of Twitter data and showed that a hierarchal cascaded pipeline of three models to label a tweet as Objective, Neutral, Positive and Negative class is better than a 4-way classifier design.
Apoorv Agarwal and Jasneet Singh Sabharwal
In Proceedings of the Workshop on Information Extraction and Entity Analytics on Social Media Data, COLING 2012, the 24th International Conference on Computational Linguistics.

Projects & Presentations

Visualizing Wikipedia using t-SNE
Presented an approach on visualizing high-dimensional data of Wikipedia using t-Distributed Stochastic Neighbor Embedding (t-SNE, specifically t-SNE using Barnes-Hut approximation). It was shown that using semantic role labels as features instead of bag of words, distinct clusters are formed for the historical human events.
Multi-Label Text Classification
Presented an approach to perform multi-label classification using Support Vector Machines which was evaluated on a subset of Large Scale Hierarchical Text Classification Challenge 4 dataset.
Presentation on Visualizing Text using t-SNE
Presented an overview on how t-SNE is useful in visualizing text, specially in Lensing Wikipedia.
Presentation on Semi-Supervised Learning
Presented an overview on various methods in semi-supervised learning at the SFU Machine Learning Reading Group.
Presentation on BRAT Rapid Annotation Tool
Presented an overview on BRAT an online environment for collaborative text annotation. The slides show the basic features available in BRAT and a demo was given on how to setup BRAT and how various types of annotations can be performed.
Inspired by Andrej Karpathy's website.