- 學位論文
以語意網技術建置的文章內容為對象的信賴評估方法之研究
Semantic Web Technology to Build the Article Content Objects of the Trust Evaluation Methods
摘要
維基百科全書的興起，讓我們見識到共筆的力量，然而也正是共筆的原因，讓我們 不免對它提出疑問，它的文章品質真的可靠嗎？會不會是三教九流胡謅亂掰？許多 研究針對維基百科全書提出了信賴評價機制，試圖從其中找出規則、檢驗文章，幫 助讀者選擇優良的文章閱讀，避開粗製濫造的文章。最常見的評估方法是使用信譽 累積機制和觀察文章的背景因素(編輯頻率、討論熱度等)，不過針對文章內容的評 價方法卻略顯匱乏，我們認為文章的語意內容應當是評價信賴最重要的線索，因此 在本篇論文裡我們提出了一個基於文章語意內容的信賴評價機制。 我們挑選使用語意網技術的維基社群作為我們的實驗對象，分別是DBpedia 跟 Freebase。語意網技術把文章內容予以結構化，因此機器能夠自動化的處理文章的語 意內容，協助我們處理信賴評價的語意計算。 我們運用 PageRank 的概念作為我們信賴評估的基礎。我們以DBpedia 的高品質 知識內容作為評分表，分析Freebase 跟DBpedia 的文章語意內容相似程度，決定 Freebase 能夠獲得多少評量分數，即Freebase 能夠獲得多少條來自DBpedia 的超連 結推薦。 我們實作了一個系統，讓使用者在瀏覽 Freebase 的條目頁面時，透過一些簡單 的操作，就能夠獲得文章的信賴評價，協助使用者判斷文章的品質，選擇優良的知 識內容閱讀。
並列摘要
The rise of Wikipedia, let us insight into the power of the co-authoring, but also because it is of the co-authoring, so that we can not help but to question it, it's the article quality is really reliable? Will not be part of everyday nonsense? Many studies for the Wikipedia proposed trust evaluation mechanism, trying to find out which rules, test articles, to help readers choose a good article to read, to avoid shoddy article. The most common assessment method is to use the credit accumulation system and the observation of the background of the article (Editor frequency, number of discussion, etc.), but the evaluation method for the content of the article is slightly scarce, we believe that the semantic content of the article should be the most important clues. Therefore, in this paper, we propose a trust evaluation mechanisms based on semantic content of the article. We choose two wiki community which use Semantic Web technology as our subjects, DBpedia and Freebase respectively. With Semantic Web technology, the article content to be structured, so the machine can handle semantic content of the article, to help us deal with the trust evaluation of semantic computing. We use the concept of PageRank as a basis for our trust evaluation method. We use DBpedia which has high quality knowledge content as the score sheet, score Freebase article with DBpedia semantic content of the degree of similarity, that is, how many hyperlinks that Freebase accepted from DBpedia as the recommendation. We implemented a system that allows users to browse Freebase entry page, through some simple operations, users can get the trust assessment of the article, help users determine the quality of the article, choose a good article to reading.
參考文獻
延伸閱讀
- Chen, J. W. (2010). 定義型文章評分量表的發展與其效度考驗研究 [master's thesis, National Tsing Hua University]. Airiti Library. https://doi.org/10.6843/NTHU.2010.00589
- 林秀芬、李國光（2006）。網路知識社群服務品質評估模式建構之研究。Electronic Commerce Studies，4(2)，211-234。https://doi.org/10.29767/ECS.200606.0004
- 藍敏杰（2006）。潛在語意分析之摘要評量系統─探討特定領域文章、關鍵詞、奇異值保留個數對建構語意空間及評量效度之影響〔碩士論文，國立臺灣師範大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0021-0712200716124994
- Krishnan, R. B., & Sakthivel, N. (2012). Development of an Efficient QoS based Web Services Compositions Mechanism for Semantic Web. Research Journal of Applied Sciences, Engineering and Technology, 4(8), 987-994. https://www.airitilibrary.com/Article/Detail?DocID=20407467-201204-201411100021-201411100021-987-994
- Lin, Y. F. (2007). A Study on Agent Communication over the Semantic Web [doctoral dissertation, National Central University]. Airiti Library. https://www.airitilibrary.com/Article/Detail?DocID=U0031-0207200917351205