1. NCCU Academic Hub
  2. 學術產出
  3. 資訊學院
  4. 會議論文
  5. Item

學術產出-會議論文

文章檢視/開啟

書目匯出

Google ScholarTM

政大圖書館

引文資訊

TAIR相關學術產出

題名 災難事件中社群媒體訊息之自動分類設計
作者 施旭峰;李蔡彥;鄭宇君;陳百齡
貢獻者 資科系
關鍵詞 災難傳播;社群媒體;鉅量資料;機器學習;社群感知
Disaster communication; Social media; Big data; Machine learnin`; Social sensors
日期 2014.05
上傳時間 7-八月-2014 18:09:27 (UTC+8)
摘要 近年來，當重大災難發生時，人們經常透過網路通訊工具傳遞災情或求救訊息，大量的資訊人力已無法負荷，如何在第一時間進行有效分類，以即時傳遞到適當的救災、協尋或資源調度單位，一直是救援單位的重要課題。本研究以台灣莫拉克風災期間的五個災難頻道的資料集為分析對象，包括地方救災中心報案紀錄、災情網站貼文、Twitter 等文字資料。經過文字處理與專家分類後，透過詞頻分布、分類結構組成、詞彙共現網絡等方式，探討不同頻道資料集之異同。進一步使用空間向量模型與機器學習的方法，建立社群媒體災難資料的自動化分類器。本研究的歸納與所發展出來的分類方式與資訊探索技術，將可用於開發更有效率的社群感知器。
In recent years, when disaster events occur, people often transfer information or distress messages through communication tools. As huge amounts of disaster information flows in, processing the data without the assistance of computational technologies becomes an increasingly challenging task. Therefore, understanding how to effectively classify information from social media, provide reliable information to disaster reaction centers, and assist policy decision-making is an important topic of discussion. In this study, we use the data collected during typhoon Morakot from five different channels, including the records of local disaster relief centers, the postings on disaster website, and tweets. After word processing and content classification by experts, we observe the difference between these datasets on the frequency distribution, classification structures, and word co-occurrence network. We further use the vector space model and machine learning method to train the classification model of social media information. We believe the techniques developed and results of the analysis can be used to design more efficient and accurate social sensors in the future.
關聯 Proceedings of International Conference on Information Management
資料類型 conference
dc.contributor 資科系en_US
dc.creator (作者) 施旭峰;李蔡彥;鄭宇君;陳百齡zh_TW
dc.date (日期) 2014.05en_US
dc.date.accessioned 7-八月-2014 18:09:27 (UTC+8)-
dc.date.available 7-八月-2014 18:09:27 (UTC+8)-
dc.date.issued (上傳時間) 7-八月-2014 18:09:27 (UTC+8)-
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/68441-
dc.description.abstract (摘要) 近年來，當重大災難發生時，人們經常透過網路通訊工具傳遞災情或求救訊息，大量的資訊人力已無法負荷，如何在第一時間進行有效分類，以即時傳遞到適當的救災、協尋或資源調度單位，一直是救援單位的重要課題。本研究以台灣莫拉克風災期間的五個災難頻道的資料集為分析對象，包括地方救災中心報案紀錄、災情網站貼文、Twitter 等文字資料。經過文字處理與專家分類後，透過詞頻分布、分類結構組成、詞彙共現網絡等方式，探討不同頻道資料集之異同。進一步使用空間向量模型與機器學習的方法，建立社群媒體災難資料的自動化分類器。本研究的歸納與所發展出來的分類方式與資訊探索技術，將可用於開發更有效率的社群感知器。en_US
dc.description.abstract (摘要) In recent years, when disaster events occur, people often transfer information or distress messages through communication tools. As huge amounts of disaster information flows in, processing the data without the assistance of computational technologies becomes an increasingly challenging task. Therefore, understanding how to effectively classify information from social media, provide reliable information to disaster reaction centers, and assist policy decision-making is an important topic of discussion. In this study, we use the data collected during typhoon Morakot from five different channels, including the records of local disaster relief centers, the postings on disaster website, and tweets. After word processing and content classification by experts, we observe the difference between these datasets on the frequency distribution, classification structures, and word co-occurrence network. We further use the vector space model and machine learning method to train the classification model of social media information. We believe the techniques developed and results of the analysis can be used to design more efficient and accurate social sensors in the future.en_US
dc.format.extent 916987 bytes-
dc.format.mimetype application/pdf-
dc.language.iso en_US-
dc.relation (關聯) Proceedings of International Conference on Information Managementen_US
dc.subject (關鍵詞) 災難傳播;社群媒體;鉅量資料;機器學習;社群感知en_US
dc.subject (關鍵詞) Disaster communication; Social media; Big data; Machine learnin`; Social sensorsen_US
dc.title (題名) 災難事件中社群媒體訊息之自動分類設計zh_TW
dc.type (資料類型) conferenceen