@GlobalTrends
![@GlobalTrends](/assets/img/projects/Auth.png)
Real Time News Scraping and Recommendation System - React | Tensorflow | NLP | News Scrapers. Monitors, scrapes and dedupes latest news from News API with MongoDB, Redis, RabbitMQ, TF-IDF. |
Built a single-page web application for users to browse news with React, Node.js, RPC, SOA, JWT.
Implemented a click event log processor which collects users’ click logs, then updates a news preference model for each user by NLP.
Clustered news using k-means and Word2Vec for recommendation. Designed and built an offline training pipeline for news text topic modeling by the use of Tensorflow, DNN.
Deployed an online classifying service for news topic modeling using the trained model.