網路新聞影響力預測：使用聯合新聞網資料為例; Predicting Influence of Online News : Using Data from Udn.com

NCUIR > School of Management at National Central University > Graduate Institute of Information Management > Electronic Thesis & Dissertation > Item 987654321/61066

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/61066

Title:	網路新聞影響力預測：使用聯合新聞網資料為例;Predicting Influence of Online News : Using Data from Udn.com
Authors:	梁孝平;Liang,Xiao-Ping
Contributors:	資訊管理學系
Keywords:	資料探勘;文字探勘;分類;Data mining;text mining;Classification
Date:	2013-07-23
Issue Date:	2013-08-22 12:11:07 (UTC+8)
Publisher:	國立中央大學
Abstract:	網際網路的發展，帶動了網路媒體的快速增加。根據財團法人台灣網路資訊中心（TWNIC）在2012年3月的調查，瀏覽新聞媒體網站﹝例如：Google News或奇摩新聞﹞、線上閱讀新聞，已成為許多民眾上網的主要目的之一。面對每天產生的龐大新聞資訊，讀者通常僅能選擇性的閱讀較重要或較令人感興趣的新聞。因此，一個新聞網站是否能正確的選擇具有影響力的新聞、並放置在適當的位置，是非常重要的議題。就本研究所知，現有線上新聞網站，尚未使用有效的演算法來自動預測一篇新聞未來的影響力，以協助網站管理人員挑選、排列網站上的新聞。因此，在本研究中，我們分析過去新聞文章的歷史資料，以建立各項預測指標。利用這些預測指標，本研究致力於在一則新的新聞發佈時，即預測該新聞一段時間之後的影響力，並分析各項左右影響力的因素。本研究的結果可用來協助新聞網站即時辨識真正重要、且能吸引讀者閱讀的新聞。 The development of Internet has driven the rapid increase in the number of online media. According to the survey conducted by Taiwan Network Information Center （TWNIC） in the March 2012, reading news on news websites, such as google news or yahoo news, has become one of the main activities people engaged online. Most readers only select the most important or interesting titles from the numerous daily news to read. Therefore, choosing influential news and placing them into a right position on news websites is an important issue. To our best knowledge, existing online news websites haven’t adopted an effective algorithm for predicting the influence of incoming news automatically to help the website managers select news and arrange the news on their website. In this paper, we analyzed the historical data of news online to propose predictive features. Using these proposed features, this study aims at predicting the influence of incoming news after a period of time and analyzing the relationship between the influence of news and the proposed features. The result of this study helps news websites identify really important and attractive news in time.
Appears in Collections:	[Graduate Institute of Information Management] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	653	View/Open

社群 sharing

Loading...