結合信任網絡，社會關係及語意分析之部落格推薦機制

Abstract

部落格為線上社群網路一個很好的範例，它包了以網頁為基礎並時常更新的網路日誌，其日誌依時間前後反轉排列而記載的順序來排列。而旁列的好友清單讓部落客能夠時常拜訪喜愛的部落格。本研究提出一個以結合信任模型、社會網路關係及文字語意分析的部落格推薦機制，並且描述此推薦機制如何應用到台灣－無名小站這個有名的部落格系統上。從實驗結果來看，我們從發展的部落格網路中發現許多的意涵並實證社會網路中重要的理論。實驗評估顯示我們提出的部落格推薦機制相當的可行並且具有潛力的。

Weblog is a good paradigm of online social network which constitutes web-based regularly updated journals with reverse chronological sequences of dated entries, usually with blogrolls on the sidebars, allowing bloggers link to favorite site which they are frequently visited. In this study we propose a blog recommendation mechanism that combines trust model, social relation, and semantic analysis and illustrates how it can be applied to a prestigious online blogging system – Wretch in Taiwan. By the results of experimental study, we found a number of implications from the Weblog network and several important theories in domain of social networking were empirically justified. The experimental evaluation reveals that the proposed recommendation mechanism is quite feasible and promising.