Reviews and Ratings Spam Detection
Submitted by Mohit Kumar (@mohitkum) on Monday, 15 June 2015
Section: Crisp Talk Technical level: Intermediate
The audience will be able to walk away with a summary view of the state-of-the-art techniques in reviews and rating spam detection and their application at Flipkart.
Reviews and ratings are a critical input into decision making across various domains for eg, restaurants reviews in Zomato/Yelp, product/seller reviews in E-Commerce. Due to their importance in decision making, reviews/ratings are targeted by spammers for promoting or demoting certain products or businesses. This talk will:
a) summarise the state-of-the-art in reviews and ratings spam detection.
b) share insights around application of these ideas at Flipkart.
Disha Makhija, Data Scientist, Flipkart
Mohit Kumar, Data Scientist, Flipkart