Shubham Chaudhary
Work
Shubham is a Senior Software Engineer in the recommendations team of
Scribd - the Netflix for books and audiobooks.
He is currently working on deep-learning embedding based item-item similarity recommendation models deployed at scale.
Previously, he worked in the Machine Learning team of Zomato.
He worked on projects involving NLP, deep learning based models for image classification & image quality
assessment, scalable duplicate detection for text documents and images, restaurant rating algorithm and user credibility.
Publications
Review Highlights: Opinion Mining on Reviews IML17, ACM
A. Kushwaha & S. Chaudhary October 2017, Liverpool United Kingdom
A hybrid model for rule selection in aspect extraction. Reduce the aspect extraction noise of syntactic rule approach by
using a multi-label classifier to select syntactic rules based on the sentence POS structure.
DOI: 10.1145/3109761.3158385
DOI: 10.1145/3109761.3158385
Bus Boarding Event Detection using Smartphone Sensors COMSNET16, IEEE
S. Chaudhary, T. Kaur, N. Aggarwal, B. Raman, D. Bansal, K.K. Ramakrishnan January 2016
A machine learning model for on body bus boarding activity detection using low power mobile device sensors -
accelerometer & GSM location.
DOI: 10.1109/COMSNETS.2016.7439930
DOI: 10.1109/COMSNETS.2016.7439930
Open Source
Shubham contributes to various projects on github and has also contributed to KDE in summer of code 2013.