博碩士論文 102423005 完整後設資料紀錄

DC 欄位 語言
DC.contributor資訊管理學系zh_TW
DC.creator葉晉昇zh_TW
DC.creatorYeh Chin-Shengen_US
dc.date.accessioned2015-7-7T07:39:07Z
dc.date.available2015-7-7T07:39:07Z
dc.date.issued2015
dc.identifier.urihttp://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=102423005
dc.contributor.department資訊管理學系zh_TW
DC.description國立中央大學zh_TW
DC.descriptionNational Central Universityen_US
dc.description.abstract隨著web2.0的興起以及網路頻寬不斷的提升與影音串流技術的進步,許多影音網站也隨之出現。目前排行第一名的影音分享網站為YouTube,它的影片上傳總數已經超過十億部影片。如此龐大的影片資料中,為了讓使用者可以快速搜尋到想要的影片,或是尋找有興趣的影片類別,因此,如何管理與分類這些巨量的影音內容成為主要的任務。 YouTube網站上除了上傳影片資料外,還包含其他的使用者產生的內容(User-generated content),如影片描述、關鍵字、影片標題、影片評論等等。而這些影片上傳者所產生的內容通常帶有主觀意見,並且融入了許多個人化的情緒和想法。YouTube在影片分類方法中並沒有考量情緒的因素,因此本文利用情緒分類將網路影片分類到指定的情緒類別中。 本文首先利用了影音網站的使用者產生的內容(User-generated content)來建立監督式的情緒分類方法,接著利用過去文字為基礎(text-base)的非監督式情緒分類方法來分類影片的情緒,最後我們提出了一個以影音網站的內容為基礎,結合過去情緒分類中的監督式與非監督式兩種方法將影音網站的影片有效的去分類到適合的情緒中。 zh_TW
dc.description.abstractWith the rise of web 2.0, Video Sharing Websites have become one of the several main sites where people spend most of their time. YouTube is one of the most popular Video Sharing Website and becomes the top most-trafficked website in the world. The numbers of video in YouTube have exceeded one billion. In order to allow user find the video which they are interested in. To manage and classify of these huge amounts of video in YouTube become a major task. In addition to the video which is uploaded by users, YouTube had many User-generated content. Such as description, keyword, title, comment of these video. The User-generated content in YouTube usually with subjective thoughts and containing personal sentiments and opinions. However the factor of sentiment was not the consideration in the study of YouTube classification in the past, so we focus on sentiment classification in YouTube. In this study, we collect YouTube data and build an ensemble model to classify the video in YouTube. We based on title, description, keyword, comment features. The ensemble model integrates results from the Supervised learning method and Unsupervised learning method. Our experiments are executed by the data mining tool and performed by a 10-fold cross-validation. Experiment results show that the performance of our ensemble model outperforms each individual classifier and the features we propose can effectively classify the video in YouTube to correct sentiment. en_US
DC.subject文字探勘zh_TW
DC.subject機器學習zh_TW
DC.subject情緒分析zh_TW
DC.subjectYouTubezh_TW
DC.subjectText miningen_US
DC.subjectMachine learningen_US
DC.subjectYouTubeen_US
DC.subjectSentiment analysisen_US
DC.titleYouTube影片之情緒分類zh_TW
dc.language.isozh-TWzh-TW
DC.type博碩士論文zh_TW
DC.typethesisen_US
DC.publisherNational Central Universityen_US

若有論文相關問題,請聯絡國立中央大學圖書館推廣服務組 TEL:(03)422-7151轉57407,或E-mail聯絡  - 隱私權政策聲明