廣播在歷史中扮演了重要的角色,聲音的發送已經在時空中傳遞了數十載。即使在現今通訊技術與資訊科技都進步神速的時代中,仍然成為大眾生活中主要接收訊息的工具。傳統的廣播媒體是以固定時間及頻率,對特定地區及範圍作播音;但經由網際網路普及,讓聽眾或網友,可突破時間及地域限制提供更方便服務選擇。早期廣播的內容都以磁帶或紙本儲存,隨著時間消逝,要再回顧曾經錄製的有聲資料是相當困難的,尢其“時間”對於保存重要的資料之完整性始終具有失真的威脅。本研究將從中央廣播電台所儲藏的珍貴聲音史料著手,經過數位典藏的方式轉換後,建立相關的描述索引鍵值,置於技術成熟之資料庫中,透過網際網路方便隨選搜尋存取及播放(On-demand)的介面,滿足社會對史實有聲資料研究參考的需求。 本研究使用IBM公司所開發的核心技術軟體,如Content Manager;昇陽公司(Sun)的J2EE,及XML等來建構系統,並以三角架構為系統的基本結構,其架構包括Library Server、Object Server、和Clients三個部份。整體運作係透過IBM Content Manager將實體的聲音資料存放至Object Server;而索引鍵值部份則置放於Library Server;最後再由使用者以瀏覽器來搜尋所要的資料,經由此一概念提供所需要的分散式處理資料。 大眾媒體一直是一種單向傳輸或是單項選擇的特性,廣大群眾只能接受,而在隨選傳播(On-Demand)的網路世界觀念中卻突破了此規範;使用者可隨意的選擇他樂於收聽的節目或頻道,甚至是收聽他們錯過和需要的節目,使用者不僅僅只是單向的接收者,甚至可自由的選擇所需的資訊。 「數位典藏」乃是這些年來,各個研究單位和學術單位共同致力於發展和努力的方向,藉由整合環境開發,將各種具紀念價值的文物、影音資料,透過數位化形式,置於技術成熟之資料庫中,靠相關的描述索引鍵值,將典藏資料呈現在網際網路上,提供分享及再利用之價值,除可保存重要的文化資產外,亦滿足社會對史實有聲資料研究參考的需求。本研究係致力於將中央廣播電台塵封的歷史聲音,藉由資訊技術作數位典藏之方法之研究,並附加資訊描述非結構化資料索引方式,作有效之隨選搜尋存取典藏資料,提供一方便之介面與機制將廣播原音重現於網友耳際,重新展現歷史有聲資料之活力與價值。
This study describes a digital contents development for Central Brocasting System, and managed multi-media contents into a well-organized database-like structure. CBS was established in 1917. For years, this audio system collects more than thousands valuable artworks. However, each work is a traditional audio type object but digitalize electronic data. For distributed and value-added reasons, finding a digitalized mechanism to manage them is what we pursue. Non-Structured Digital Data (NDD) is referring the digital data, which is using electronic mechanisms to store data but formatted Traditionally, structured digital data, such as a database, can be accessed by using SQL (Structure Query Language) to find useful information. However, retrieving information from NDD is hard since it needs more fundamental technologies. Various technologies were developed and mature enough to improve information retrieval. Full text search, OCR (Optical Character Recognition), and QBIC (Query by Image Content) are good examples. It is the time to standardize NDD’s application under a normalized management. The purpose of this research implements a prototype, which integrate above technologies to retrieve information from NDD and then output the result by using the Web browser as a platform. We will focus on available indexing techniques and backend NDD, which combined with Web applications development technologies. For improving the flexibility and scalability of this prototype, we plan to use object-oriented analysis technique, such as UML (Unified Modeling Language), to make them as components. The ideal system will be based on multi-tier Client/Server architecture that should be platforms free. Another hard work is displaying NDD content in the Web browser. We will base on XML (Extended Markup Language) and XSL (XML Style sheet Language) to replace HTML as the content resources. The benefit of this research is to find a reference prototype that can improve existing problems on the Web contents. For example, to get Web contents is by linking individual files and then displaying them. However, when NDD is going large, current retrieving technique is inefficiency. Since the prototype is based on a new architecture for NDD, it causes a change of the Web content from directories/files oriented to database oriented.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。