Building a Tag Map for Recommendations in Microblogging

In its brief history of about two decades, the Web has evolved from a technical framework for information dissemination to more of an enabler of social interactions among its users. This weaves a huge virtual social network for users. Finding friends and targeting useful information are great challenges in such a complicated social network. Different with traditional content-based and collaborative information filtering techniques without considering social connections among users, we propose a naive recommendation approach utilizing user relationships to find friends and hot topics in social media. This is achieved by weaving a tag map from user preferences. Our experiments based on Microblogging data from Sina.com show a promising result in social media recommendation.