Hafeezul RahmanGraduate Student
Language Technologies Institute
School of Computer Science
Carnegie Mellon University
Email: hmohamma@cs.cmu.edu
Publications
Invited Talk
Flipkart India Private Limited; Bangalore, India; August 17, 2018.
Research Internships
End to End Data Provenance Capture
May 2016 - July 2016
Harvard John A. Paulson School Of Engineering And Applied Sciences, Cambridge, MA, USA
As a research intern, I worked on developing a software tool that would enable tracking the provenance of every entity in an R script, directly at runtime. Also provided alongwith, is a visualization extension that enables viewing the data provenance graphs which could be used for debugging R scripts and compare different R script execution flows.
Practical Anonymizing Data Analytics
May 2015 - July 2015
Max Planck Institute for Software Systems, Kaiserslautern, Germany
Anonymization of information(specifically, personally identifiable information) is very important and a serious concern, yet it comes at a cost. It cannot be used for data analytics to the fullest extent possible. I worked on developing data anonymization techniques to deploy a practical anonymizing data analytics system that tries to provide rich analytics while at the same time, ensures user-anonymity. The technology underpinning the system includes trusted computing, query sand-boxing, and anonymization mechanisms, all built on scalable Map-Reduce framework coupled with PostgreSQL.
Data Mining & Machine Learning in Traffic Prediction
May 2014 - July 2014
University of Southern California, Los Angeles, USA
Traffic impact prediction refers to being able to predict in real time, short and long term traffic due to impact in a road segment. I worked on designing the implementing the Co-operative Contextual Baandit Algorithm in such a framework to study the traffic conditions, when an accident(or any other traffic impact like roadblocks, etc) happens on the LA streets.