| 摘要: | 隨著網路快速的發展，社群媒體平台成了人們生活中的一部分，如 Facebook,
YouTube 等平台的壯大，新興媒體也成為現代的主流。從人們只上網搜尋資料的 Web 1.0
時代，轉向人們開始不斷在網路上發佈作品的 Web 2.0，然而網路改變生活的同時也帶
來新問題，人們沒有意識到網路擁有權，大數據公司將用戶的資料進行銷售。
近年來，Web 3.0 開始發展，資訊擁有權成為重點討論的一部分，為了降低資訊集中
於少數者手中，去中心化技術的應用開始崛起；然而，現今主流網路服務架構仍為 client server 架構，如何無縫地轉變為以用戶端為中心 (user-centric) 的研究議題仍持續進行，
探討這些新興應用的研究與開發，造就了許多新的分散式應用型態與服務。
本研究關注的議題為重構新型態網路服務架構以降低資訊過於集中，同時也考慮降
低開發成本與無縫接軌務實面。換言之，於傳統 client-server 架構下，使用者普遍使用
瀏覽器連接至特定網頁伺服器，進行關鍵字搜尋以取得感興趣之資訊。本研究欲透過
object management group (OMG) 之 data distribution service (DDS) 技術，維持使用者已
孰悉之瀏覽器介面，透過 DDS 代理者 (Agent) 所提供發布 (Publish，Pub) 與訂閱
(Subscribe，Sub) 能力，能更便捷取得更豐富且完整資訊，資訊來源不再受限於特定伺
服器，任何資訊擁有者皆是提供者。進一步說，使用者關注感興趣之關鍵字，透過代理
者扮演 Publisher 於一特定主題 (Topic) 上發佈 (Pub) 關鍵字，現有 Google, Bing, Baidu
等主流搜尋引擎扮演 Subscriber 透過訂閱 (Sub) 此特定主題，同時取得使用者所輸入之
關鍵字，並將搜尋結果以相同 Pub/Sub 方式回傳給代理者，經過資訊彙總整理後，回傳
給使用者瀏覽器顯示。從使用者角度看，進行關鍵字搜尋後，便可以獲得所有訂閱主題
者所提供之相關內容。
當多數使用者與資訊提供者皆使用此架構時，使用者在不改變使用習慣前提下，可
ii
輕易且快速取得所有資訊提供者之內容。也因此，網路服務成為以用戶為中心之架構，
打破過去長期資訊壟斷的現象。;As the internet is rapidly deployed, social media has become a part of people’s lives. The
well-known social media platforms, like Facebook and YouTube, make people search for
information from the server and publish their idea via the server. Such internet service is the
so-called social web or Web 2.0. Daily life has been changing by the internet, but new problems
emerge, one of which is big data companies benefit from personal information, but internet
information providers do not realize that they shall have ownership.
The development and discussion about internet information ownership have become an
important part of Web 3.0. To decrease the monopoly of centralization, decentralized network
technologies have received much attention. Nevertheless, the client-server-based service model
is still the major service model on the internet. How to smoothly transform the service model
from client-server to user-centric is ongoing. Various contemporary and distributed services
and applications have been proposed and developed in recent years.
This study aims to propose a new service model which is able to decrease the degree of
information centralization, lower the development cost and seamlessly transfer. Under the
traditional client-server service model, users often utilize the browser to connect to a designated
web server and input the keywords for searching the information he/she is interested in. This
study proposes to utilize the object management group (OMG) data distribution service (DDS)
technology in order not to change the familiar browser of users, but users can retrieve much
more information from all information providers via the publish/subscribe ability of the DDS
agent which is developed in this work. Specifically, the user still uses a browser to input
keywords, and the developed DDS agent will play the role of publisher to publish the keywords
on a specific topic. Then, those well-known search engines, such as Google, Bing, Baidu, etc.,
will play the role of subscribers to receive the keywords by means of a subscription of the topic.
iv
After then, each search engine will reply its search result to the DDS agent independently via
the same Pub/Sub model. Upon the DDS agent collects the corresponding replies, it returns the
re-organized web page to the user’s browser for rendering. From the user’s point of view, the
user obtains all of the results from those subscriptions after it uploads a keyword once.
As the majority of users and information providers participate in the proposed DDS-based
service model, users can easily and fast access information. Eventually, the network service
model is transformed from a network-centric client-server model to a user-centric peer-2-peer
model, and the monopoly of centralization in the past decades is resolved as desired. |
