Built a keyword extraction tool/API to assist small businesses to detect counterfeit products on Korean E-commerce websites
Extracts keywords from e-commerce items in different categories and provides keywords/price distribution information with user’s feedback input;
Built a web-crawler customized for G*CCI website in selenium; extracts and organizes all G*cci products’ data in a pre-defined format which is later inputted into keyword API
Embedded to Faikerz’s main model as a filter to screen ~40% items from the training dataset which increased overall detection accuracy by 30% and computation efficiency by 40%
- StacksPython, Flask, Selenium, MongoDB