整合深度學習技術與SOR理論之資訊情緒傳遞性探索：新聞生成特質與資訊情感傳播行為;Integrating Deep Learning Technique And SOR Theory to Explore The Transmission of Information Emotion: News Generation Characteristics And Information Emotion Dissemination Behavior

NCUIR > School of Management at National Central University > Graduate Institute of Information Management > Electronic Thesis & Dissertation > Item 987654321/89939

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/89939

Title:	整合深度學習技術與SOR理論之資訊情緒傳遞性探索：新聞生成特質與資訊情感傳播行為;Integrating Deep Learning Technique And SOR Theory to Explore The Transmission of Information Emotion: News Generation Characteristics And Information Emotion Dissemination Behavior
Authors:	董欣澄;Dong, Xin-Cheng
Contributors:	資訊管理學系
Keywords:	BERT;情緒分析;SOR model;新聞媒體;社交媒體新聞分享;BERT;Sentiment Analysis;SOR model;news media;social media news sharing
Date:	2022-09-26
Issue Date:	2022-10-04 12:05:17 (UTC+8)
Publisher:	國立中央大學
Abstract:	近年來社交媒體興起，閱聽者消費新聞的方式發生變化，傳統新聞媒體也朝向網路發展，新聞媒體間的競爭日漸加劇，競爭的環境衍伸出許多問題，新聞報導理應中立且客觀，然而在資訊爆炸的時代下，新聞媒體傾向在撰寫標題時加入了許多會激發閱聽者情緒的詞彙，利用誇大且聳動的新聞標題吸引閱聽人點擊新聞以賺取更多的廣告收入，新聞媒體此種寫作手法違背自身所應遵守的原則，最終失去民眾對於新聞媒體的信任與尊重，同時社群媒體將新聞傳播得更快更廣，用戶更容易暴露於負面的新聞環境中，長期下來除了誤導閱聽人外更會影響其心理健康和情緒。因此，本研究目的為探討新聞文本內容的生成特質對於資訊情感傳播的連鎖效應。本研究蒐集台灣主流媒體新聞，並訓練BERT模型分類出影響閱聽人情緒的新聞標題，另外透過以刺激-有機體-反應(Stimuli-Organism-Response, SOR)理論進行統計分析，藉此了解網路新聞特質對於閱聽人於社群媒體傳播影響力和刺激反應。結果顯示，新聞標題情緒會顯著影響閱聽人傳播行為和傳播後的情緒反應。首次將SOR理論引入社交媒體新聞分享，除了為未來相關研究提供基礎，也提供新聞媒體與閱聽人了解數位新聞環境所面臨的問題。;In recent years, with the rise of social media, the way readers consume news has changed, and traditional news media have also developed online. The competition among news media has intensified. Many problems have arisen from the competitive environment. News reporting should be neutral and objective. In the era of information explosion, news media tend to add a lot of words that will arouse the emotions of readers when writing headlines, and use exaggerated and sensational news headlines to attract readers to click on news to earn more advertising revenue, this kind of writing style violates the principles that one should abide by, and ultimately loses the public’s trust and respect for the news media. At the same time, social media spreads news faster and more widely, and users are more likely to be exposed to a negative news environment. Misleading readers can even affect their mental health and mood. Therefore, the purpose of this study is to explore the chain effect of the generative characteristics of news text content on the emotional dissemination of information. This study collects news from mainstream media in Taiwan, and trains the BERT model to classify news headlines that affect readers′ emotions. In addition, it conducts statistical analysis based on the Stimuli-Organism-Response(SOR) theory to understand the online News characteristics are influential and stimulative for audiences to spread on social media. The results show that news headline emotions can significantly affect readers′ dissemination behavior and emotional responses after dissemination. The introduction of SOR theory into social media news sharing for the first time not only provides a basis for future related research, but also provides news media and readers with an understanding of the problems faced by the digital news environment.
Appears in Collections:	[Graduate Institute of Information Management] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	186	View/Open

社群 sharing

Loading...