CONTENTS RECOMMENDATION SYSTEM AND METHOD BASED ON SOCIAL NETWORK

摘要

PURPOSE: A social network-based content recommending system and a content recommending method thereof are provided to utilize internet resources' tags and user characteristics beyond the existing simple keyword matching. CONSTITUTION: A content recommending system (10) collects the characteristics (C) and the tags (T) of users who are the members of a community from an SNS repository (20). The content recommending system integrates the collected characteristics and tags into three matrices (Nac, Mat, Qact) and composes a graph (S10). Using the graph, the center of each node is calculated through a social network analysis procedure (S20). The content recommending system extracts a hot tag and general characteristics using the calculated center and selects content which includes the hot tag for recommendation to a user (S30). [Reference numerals] (20) SNS repository; (c) Characteristics; (CC) Common characteristics; (HT) Hot tag; (S10) Material integration; (S30) Content recommendation; (t) Tag