Full metadata record
DC FieldValueLanguage
dc.contributor.author林俊碩en_US
dc.contributor.authorLin, Chun-Shuoen_US
dc.contributor.author彭文志en_US
dc.contributor.authorPeng, Wen-Chihen_US
dc.date.accessioned2014-12-12T01:43:15Z-
dc.date.available2014-12-12T01:43:15Z-
dc.date.issued2010en_US
dc.identifier.urihttp://140.113.39.130/cdrfb3/record/nctu/#GT079755506en_US
dc.identifier.urihttp://hdl.handle.net/11536/45852-
dc.description.abstract在Web 2.0的時代,微網誌在網路中不斷的傳遞大量的資訊,其儼然已成為一種相當重要的資 訊傳播型式。藉由手持式裝置的協助,微網誌可以發布更具有時空敏銳度的訊息。由於微網誌 的訊息量相當的龐大,現有的微網誌搜索引擎,如TwitterSearch,已經不能滿足我們的需求 以總結出某個時間/空間上所發生的事件。在給定地點及時間後,這篇論文的目的是在微網誌 的資料中,偵測在該時間/地點所發生的事件(spatio-temporal event)。一個spatio-temporal event必須滿足以下三個標準:一)空間上的集中度;二)時間上的集中度;三)普及度。為了 達成這個目標,我們提出了一個架構,包涵了4個階段:(1) profile construction (2) feature extraction(3)event summary detection(4)ranking events。在本論文中,我們提出STF (Spatio-Temporal Focus)去計算一個Feature的獨特性。此外,我們結合了STF與文件重 疊度,將一個事件從event seed開始做擴展。為了取出k個可能的 event summary,我們提出 了一個有效率的top-k soft clustering algorithm。在實驗的部份,我們使用來自於Twitter 的真實 資料去驗證我們所提出的架構。zh_TW
dc.description.abstractIn the age of Web 2.0, microblogs have become an important form of media, spreading information in the web. With the help of mobile devices, microblogs can spread the information with greater spatiotemporal sensitivity. With the overwhelming number of microblog messages, existing microblog search engines, such as twitter search, could not satisfy our need to summarize events within a location and time interval. This work aims at detecting spatio-temporal events in microblogs with an pair \{location, time\} as the input. A spatio-temporal event must satisfy the following 3 criteria: i) spatial focus; ii) temporal focus; iii) popularity. To achieve this goal, we propose a framework with 4 phases: (1) profile construction; (2) feature extraction; (3) event summary detection; (4) ranking events. In this paper, we propose STF(Spatio-Temporal Focus) value to evaluate the distinctiveness of a feature. Furthermore, we combine STF value with document overlap to expand an event cluster from an event seed. To extract the top-k possible event summaries, an efficient top-k soft clustering algorithm is proposed in this paper. In the experiments, we use a real data set from Twitter to verify our proposed framework.en_US
dc.language.isoen_USen_US
dc.subject微網誌zh_TW
dc.subject事件偵測zh_TW
dc.subjectmicroblogen_US
dc.subjectevent detectionen_US
dc.title對微網誌進行時空階層架構的事件偵測機制zh_TW
dc.titleSpatio-Temporal Hierarchical Event Detection in Microblogsen_US
dc.typeThesisen_US
dc.contributor.department資訊科學與工程研究所zh_TW
Appears in Collections:Thesis