以內容標籤的關聯性為基礎之社群分享平台的設計與實作

Abstract

本研究目的是設計與實作出以標籤關聯性為基礎的社群分享平台，隨著行動裝置的興起，越來越多人利用行動裝置來紀錄生活的點點滴滴，社群網路服務的興起讓照片的分享更加方便，然而隨著照片的數量急遽增加，要把這些內容（Content）做整理與分享，變成一件很麻煩的事情。 為了達到方便的內容（Content）管理與存取，在之前的研究中已經提出以內容為中心的家用網路平台 (Content-Centric Platform for Home Networks，簡稱CCP)，使用者透過CCP可以利用標籤(Tag)和標籤間的關聯性(Tag Association)來管理並搜尋內容，然而CCP一開始的設計只針對在家庭網路中的內容分享。 在本論文提出以內容標籤的關聯性為基礎之社群分享平台(Tag Association-based Content Sharing Platform for Social Groups,簡稱TA-CSP)，與CCP相似，TA-CSP一樣提供使用者可以用標籤和標籤之間的關聯性管理並分享內容，另外，在TA-CSP中可以讓使用者彈性地選擇分群，可以從社群網路中動態地選取要分享的對象，再者TA-CSP可以動態整合感興趣的朋友們所分享的內容，並將其標籤間關聯性重組，而不需要重新分析內容的標籤，最後TA-CSP還提供在個人設備上利用標籤關聯性做內容管理的功能，讓每台設備可以建立自己內容的標籤關聯性，讓使用者更方便地管理設備上的內容。 本論文提出一個系統雛型來驗證TA-CSP的實用性，由結果可得知，TA-CSP適用於家庭網路及社群團體中，可以達到以標籤關聯性為基礎的內容管理及分享機制，期許能給使用者一個更簡單瀏覽、管理和分享內容的平台系統。

The aim of this research is to design and implement a tag association-based content sharing platform for social groups. With the popularity of mobile devices, people can now take pictures anytime, anywhere. Furthermore, the prevalence of social networking services further makes it ever convenience for people to share their pictures. However, as the number of pictures increases it is becoming more and more troublesome to manage (searching/browsing) pictures to share with others or shared by others. To facilitate content management and fast content retrieval, a Content-Centric Platform (CCP) has previously been proposed. With CCP, users may manage and search contents according using the tags (or attributes) of contents and tag associations among contents. However, CCP is a platform that designed specifically for content sharing among a fixed group in a home network environment. In this thesis, we propose a Tag Association-based Content Sharing Platform (TA-CSP) for Social Groups. Similar to CCP, TA-CSP allows user to manage contents according to tags of contents and tag associations among contents. Nevertheless, it is more flexible in group management for content sharing. It allows users to select friends dynamically from social networks and publish contents to the group of the selected friends. Furthermore, TA-CSP can aggregate dynamically the tags the contents published by interested friends and establish the association of tags among the contents incrementally without parsing the tags of all shared contents. Last but not least, TA-CSP also supports tag association-based management of local contents by having each device to establish the tag associations of local contents first. Finally, we have developed a prototype system to verify the effectiveness of the TA-CSP. Experimental results show that TA-CSP can achieve customized tag association-based content management and sharing for social groups.