| DC 欄位 |
值 |
語言 |
| DC.contributor | 企業管理學系 | zh_TW |
| DC.creator | 魏天昊 | zh_TW |
| DC.creator | Tien-Hao Wei | en_US |
| dc.date.accessioned | 2016-6-20T07:39:07Z | |
| dc.date.available | 2016-6-20T07:39:07Z | |
| dc.date.issued | 2016 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:444/thesis/view_etd.asp?URN=103421043 | |
| dc.contributor.department | 企業管理學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 中文摘要
特殊事件的發生和人類情緒之間有著緊密的關聯性，情緒字詞是人們用來有效描述特殊事件的重要媒介（Kolya et al., 2011），社群網站的使用者情緒會因為現實世界中特殊事件的發生而產生明顯差異（Kramer, 2010; Choudhury et al., 2008）。透過社群網站所蘊含的大量情緒成分，本研究針對事件情緒、發文情緒和產品推薦情緒進行探討，並分析如何利用情緒變數來引起使用者對於產品推薦的點擊興趣。
為了解情緒如何影響推薦效果，本研究使用微網誌平台Plurk配合卓淑玲、陳學志、鄭昭明（2013）的中文情緒詞庫，以情緒關鍵字為根據來判斷使用者的發文情緒，同時利用情緒字詞為產品推薦的依據，並且進一步透過T檢定、one-way Anova、two-way Anova探討不同情緒變數對於產品推薦效果的影響。
由實驗結果發現，在正向事件的影響下，微網誌使用者的負向情緒發文會比正向發文更能夠提升產品推薦效果。在帶有不同情緒字詞的產品推薦中，正向情緒推薦比起負向和中立推薦更能達到推薦效果的提升。除此之外，在負向發文底下予以正向情緒的產品推薦能帶來最佳的產品推薦效果。另一方面，在負向事件的影響下，微網誌使用者的負向情緒發文會比正向發文更能夠提升產品推薦效果。在帶有不同情緒字詞的產品推薦中，負向情緒推薦比起正向和中立推薦更能達到推薦效果的提升。除此之外，在負向發文底下予以負向情緒的產品推薦能夠帶來最佳的產品推薦效果。
藉由本研究的結果，能有效幫助服務或產品供應商在特殊事件發生期間進行社群行銷，以利用情緒變數的方式有效吸引社群網站用戶對廣告的點擊興趣，進而達到提升產品推薦效果的目標。 | zh_TW |
| dc.description.abstract | Sentiment and Event are two abstract entities closely coupled with each other. Sentiment terms are important medium that people can describe events（Kolya et al., 2011）. The occurring of events will bring significant difference in user sentiment of social media（Kramer, 2010; Choudhury et al., 2008）.Through the sentiment component of social media, our research focus on the sentiment of event, user posts and product recommendation. We analyze sentiment variables and discuss how to arouse user attention in order to improve the effectiveness of social marketing.
In order to realize the influence of sentiment, we apply the microblog platform Plurk in line with Chinese sentiment lexicon provided by Jhuo et al. (2013). Based on sentiment lexicon, we extract and classify user sentiment according to sentiment keywords emerge in posts. Apart from this, we use the lexicon as foundation of product recommendation to observe the influence of sentiment words.
Under the influence of positive event, our results show that negative posts and positive recommendation are more effective to product recommendation. Besides, negative recommendation under the positive posts will bring the most effective outcome. On the other hand, under the influence of negative event, our results show that negative posts and negative recommendation are more effective to product recommendation. Besides, negative recommendation under the negative posts will bring the most effective outcome.
Through the experiment results, service or product suppliers can use sentiment variables on social marketing during event period to achieve better quality of product recommendation. | en_US |
| DC.subject | 特殊事件 | zh_TW |
| DC.subject | 情緒 | zh_TW |
| DC.subject | 微網誌 | zh_TW |
| DC.subject | Plurk | zh_TW |
| DC.subject | 產品推薦 | zh_TW |
| DC.subject | event | en_US |
| DC.subject | sentiment | en_US |
| DC.subject | microblog | en_US |
| DC.subject | Plurk | en_US |
| DC.subject | product recommendation | en_US |
| DC.title | 特殊事件發生期間產生之情緒效應對微網誌產品推薦效果之影響 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | The Effect of Event Sentiment on Product Recommendation in a Microblog Platform | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |