隨著網際網路的蓬勃發展，網路上的內容多采多姿，從早期的文字、圖片、到現在的多媒體 影音，已成為現代人不可或缺的娛樂之一。受惠於串流技術的廣泛應用，越來越多人選擇透過聯 網電視觀看影音節目’而近年來STB (Set-Top Box)的強大發展潛力’也讓線上影音內容朝向電視 發展。用電視上網觀看網路影音的概念已漸漸普及也然而，面對網路上如此包山包海巨量的影音 內容，如何快速有效的整理出消費者感興趣的內容也成為一項挑戰。本技術提出的影音搜尋與關 聯分析技術’整理各類影音大站之UGC (User-Generated Content)內容’成為多個完整且有效的節目 頻道。本論文所提方法具2大特色，包括1.節目集成關聯頻道及2.失效節目替換機制，解決現有影 音平台與搜尋引擎的不足。換句話說，當各影音平台內容越豐富，頻道的更新與互補效果越顯著。 為驗證本技術是否符合使用者需要銷我們將系統平台實際上線供民眾試用，從2012年6月至截稿 前續已超過2萬2千名用戶註冊登入使用，證明本平台確實對網路族群有一定吸引力。未來我們將 持續增益系統功能，並整合社群行銷以提升用戶黏著度。

Asthe Interne t developed rapidly, information containing text, pictures and videos has become extremely various and necessary.Because of the upswing of streaming technology, more and more people choose to watch video programs through Connected-TV. In recent years, online videos have become more like TV programs because of the manufacture of STB (Set-Top Box), and the idea of watching videos on TV through Internet has become another new Therefore, how to efficiently sort out those videos which
interest ^S，3 consumers most when facing the immense informationhas been abig chalienge. We propose searching and analysis technQlogy for videos, whioh classify UGC (User-Generated Cmtent) in
well-kAown video to several complete and reachable program channels. There are two main
features in our proposed method, which concludes 1) “Gathering Relational Programs h Channel，1 and 2) “Substituting Useless Programs Automatical|y，，，to enhance insignificance of internet video platforms and search engine. In other words, the various lhe video contents are, the obviousthe mutuai effects between cross-platforms will reveials. In order to verify whether this technology fits in with users' needs，we realize it to be 3n on-line platform for tricol. From June 2012，over 22，000 users have logged-jn and use our platform， which shows tn；rt this platform is appeming to certain hternet users. In the futu「e，we'll continuously enhance whe system with integratjon of social functionalities to increase the Goherence with the users.