以社會性標籤為基礎的擴充搜尋技術支援影音分享網站中之影片檢索

A Social-Tag-Based Query Expansion Approach for Supporting Video Retrieval in Video Sharing Websites

李彥賢(Yen-Hsien Lee) ； 楊錦生(Chin-Sheng Yang) ； 廖國堯(Guo-Yauo Liau)

《資訊管理學報》 19卷3期 (2012/07) Pp. 533-565

https://doi.org/10.6382/JIM.201207.0533

摘要

WEB 2.0的概念興起促使網路使用者從資訊接收者的角色轉變成資訊生產者，並透過適當的網路平台與其他網路使用者進行資訊分享與互動。近年來，隨著網路與資訊科技的快速發展，已使網際網路中分享的資訊媒體從過去單純的文字內容逐步演進到以影音多媒體為主流，並發展出許多影音分享網站。然而，為方便使用者在龐大的分享影音資料中進行搜尋，此類網站往往會提供影片搜尋機制。不若以文字描述為主要內容的傳統網站，影音分享網站每則影音檔案都只有少量的描繪資料可供進行關鍵字比對，容易在比對時造成字串錯配（Word Mismatch）的問題，進而影響影音檢索的效能。本研究提出以社會性標籤為基礎的擴充搜尋技術（STBQE），先藉由局部分析方法取得初步的影片搜尋結果，再試圖利用使用者分享影片時，針對影片內容所制定的描繪標籤，來進行搜尋字串擴充以及後續搜尋結果排序。本研究最後利用實證評估方式來比較現有影音分享網站的搜尋機制與本研究提出STBQE的差異，而結果發現本研究提出擴充搜尋技術能在特定情境之下改善現有方法的搜尋效能。

關鍵字

搜尋字串擴充 ； 影音檢索 ； 社會性標籤 ； 情境感知 ； 搜尋引擎

並列摘要

The advance of Internet and information technologies has turned the information sharing on the Web from the form of textual content into that of multimedia and fostered the development of video sharing websites. To ease the access of the large number of online videos, the sharing websites usually provide web surfers the mechanism for searching their targeted videos. However, traditional searching approaches that focus on the analysis of textual contents may not effectively support the search for the online videos, which generally contain a few annotations given by the providers. As a result, keyword search has been one of the familiar approaches to the video retrieval, although it may suffer from the word mismatch problem. In this study, we intend to improve the effectiveness on retrieving the online video by proposing the Social-Tag-Based Query Expansion (STBQE) approach. To identify from the initial search results the context words relevant to the user's query, our approach can use which to expand the search and well rank the search results. Our evaluation results also suggest that the effectiveness of our proposed STBQE approach is comparable to the performance benchmark, YouTube.com, under some evaluation scenario.

並列關鍵字

query expansion ； video retrieval ； social tagging ； context aware ； search engine

參考文獻

Attar, R.,Fraenkel, A. S.(1977).Local feedback in full-text retrieval systems.Journal of the ACM.24(3),397-417.

Au Yeung, C. M.,Gibbins, N.,Shadbolt, N.(2007).Understanding the semantics of anbiguous tags in folksonomies.Proceedings of the International Workshop on Emergent Semantics and Ontology Evolution.(Proceedings of the International Workshop on Emergent Semantics and Ontology Evolution).:

Baeza-Yates, R.,Ribeiro-Neto, B.(1999).Modern Information Retrieval.New York, NY:Addison Wesley.

Ballan, L.,Bertini, M.,Bimbo, A. D.,Meoni, M.,Serra, G.(2010).Tag suggestion and localization in user-generated videos based on social knowledge.Proceedings of Second ACM SIGMM Workshop on Social Media.(Proceedings of Second ACM SIGMM Workshop on Social Media).:

Billhardt, H.,Borrajo, D.,Maojo, V.(2002).A context vector model for infirmation retrieval.Journal of the American Society for Information Science and Technology.53(3),236-249.