Machine Learning on Blog Service

講者:施晨揚 Kent / PIXNET
地點:3F – 第一會議室
講題:Machine Learning on Blog Service


Pixnet is the most popular blog service at Taiwan. More than 400,000 articles have been published by user and 28,000,000 page views at everyday but only 2% of users have logged-in
at At now the recommendation and advertisement could not provide the personalization service for the 98% of user who did not login to blog service.
So need a solution to predict the others 98% of user’s profile.
Such as we use click log and article analysis to build related model that could describe user's demographic. Those profile analysis could help
to provide more better service to users.
So we have two major research topic as below.
1. User Demographic Detection
2. Blog Article Analysis


熱愛 Machine Learning & Data Analytics