關鍵詞擷取與焦點新聞偵測

Keyterms Extraction and Hot News Detection

簡永鑫(Yung-Shin Chien)

指導教授 ： 吳昇

國立中正大學/工學院/資訊工程學系/碩士(2013年)

摘要

關鍵詞擷取技術對於文章知識的重點擷取扮演非常重要的角色。在資訊檢索領域中，對於資訊重點進行擷取也有助於提昇檢索結果的準確度。如何提昇關鍵詞擷取的準確度及關鍵詞擷取的應用也是我們相當有興趣的範圍。

本論文實作了一個關鍵詞擷取的方法，使用TF-IDF演算法計算關鍵詞的權重，並利用引號區塊詞彙擷取及詞彙擴張法產生新詞提昇關鍵詞擷取的準確度，最後將關鍵詞擷取技術應用在焦點新聞偵測的方法。

關鍵字

焦點新聞偵測 ； 關鍵詞擷取

並列摘要

Keyterm extraction is an important technology of Information Retrieval. For search engine, it could enhance search results’ precision by extracting keyterms of information. We are interested in how to enhance the precision of keyterm extraction, and applications of keyterm extraction.

In this thesis, we implemented a method of keyterm extraction. We use the TF-IDF algorithm to evaluate the weight of keyterms, and then improved the precision of keyterm extraction by quotes block extraction and term expansion. Finally, we apply the keyterm extraction method to detect hot news in daily news pool.

並列關鍵字

Hot News Detection ； TF-IDF ； Keyterm Extraction

參考文獻

[11] Christopher D. Manning; Prabhakar Raghavan; Hinrich Schütze (2008). Introduction to Information Retrieval. Cambridge University Press. p. 118.

[1] N-gram, http://en.wikipedia.org/wiki/N-gram

[2] UTF-8, https://en.wikipedia.org/wiki/UTF-8

[3] libcurl - the multiprotocol file transfer library, http://curl.haxx.se/libcurl/

[4] WordNet, http://wordnet.princeton.edu/