Title: Data Science@Flipkart
Speaker: Mohit Kumar, Flipkart
Abstract:
Flipkart, founded by IIT D alums, has built a reliable and scalable
technology framework which is enabling eCommerce in India. We believe
that in order to provide awesome experience to our customer, its not
sufficient for our technology system to just enable decision making. We
need to build systems which are intelligent and can take decision
themselves. Systems which can take complex and fuzzy information and
translate them into timely and accurate decisions aka Systemic
Intelligence.
In this talk, we will highlight a broad set of problem statements in the
areas of personalization, recommendation, pricing, demand forecasting,
supply chain optimization, reviews mining/summarization, social media
sentiment analysis, fraud detection that require application of Machine
Learning, NLP, Optimization techniques. The application of these
techniques enables the Systemic Intelligence. We will describe the usage
of ML/NLP/Stats techniques particularly in the problem area of Fraud
detection/prevention. There are several types of fraud that are present
in an e-commerce environment like credit card fraud, reseller fraud
(bulk buying of products with the intention to re-sell), reviews and
ratings fraud (with the intention of boosting a seller/product ratings
or pulling down competition). We will describe how different techniques
are applicable in solving different types of fraud.
Bio:
Mohit is a Data Scientist with Flipkart. His research has been in the
areas of active learning, machine learning, data mining, speech and
summarization. Prior to joining Flipkart, he has worked with Accenture
Technology Labs, Chicago for five and half years in the Analytics
research group and with IBM India Research Lab, Delhi for 2 years. He
has done his MS + PhD in Computer Science from Carnegie Mellon
University and B.Tech in Computer Science and Engineering from IIT
Kharagpur.