以自然語言處理分析社群網路願望之研究

Detecting Chinese Wish Messages in Social Media and Categorizing into Knowledge Base

張衡(George Chang)

指導教授 ： 許永真

國立臺灣大學/電機資訊學院/資訊網路與多媒體研究所/碩士(2013年)

https://doi.org/10.6342/NTU.2013.11038

摘要

夢想與希望自古以來象徵著人類的價值與前進的動力。在網路社群(Social Media)與微網誌盛行之時，網路的隔世造就世人輕易於雲端坦白露念，因此期許與願望不再受限於噴泉、宗教神氏或流星殞落之時。收集分析為網誌之中的研究願望，不但能從中發探究商場產品的趨勢與潛在市場，也能挖掘特殊需求並提供解決方案，受惠企業、百姓與弱勢族群。藉由分析一個活躍於港澳台區域的 Linkwish 行動願望社群網站。我們得以歸納了解願望之特性與內容，並以支援向量機(Support Vector Machine)搭配多種語言特徵作為依據，偵測網誌是否為願望，並藉由圖樣分析取得其目標資訊，終將願望分類至知識庫(Knowledge Base)做為具有認知意義的分類。以便用於檢索與統計。本篇論文使用語言特徵能提升願望偵測準確達0.95 AUC，對於精簡明確的願望能準確分析出願望目標資訊，並分類至知識庫。

關鍵字

自然語言處理 ； 社群網路 ； 願望

並列摘要

People have wishes and sometimes share their wishes in social media, hoping to get greetings or to find partners with the same wishes. By collecting and analyzing those wishes, we may find out not only the trend of common wishes, but also the needs of individuals. This paper presents a preliminary study of Chinese wish analysis. We provide analysis on the data from Linkwish, which is a micro social network for wish sharing with users mainly from Taiwan, Hong Kong, and Macao. Then, we use SVM with various types of features to classify these messages as wish or not, extract wish target information, and categorized wish into knowledge base. Our experimental results show that some features in wish detector can achieve average areas under precision-recall curves higher than 0.95 in 10-fold cross validation, And extract target, link into knowledge base from simple wishes.

並列關鍵字

linkwish ； Social Media ； NLP

參考文獻

[7] G. S. Speer. Oral and written wishes of rural and city school children. 10:151––155,1939.

[8] Y.-F. Tsai and K.-J. Chen. Reliable and cost-e↵ective pos-tagging. International Journal of Computational Linguistics and Chinese Language Processing, 9:83–96, 2004.

[1] E. Agichtein and L. Gravano. Snowball:extracting relations from large plain-text col- lections. In Proceedings of the 5th ACM International Conference on Digital Libraries, pages 85–94, 2000.

[2] C.-C. Chang and C.-J. Lin. LIBSVM: A library for support vector machines. ACM Transactions on Intelligent Systems and Technology, 2:27:1–27:27, 2011. Software available at http://www.csie.ntu.edu.tw/~cjlin/libsvm.

[3] X. Ding, B. Liu, and P. S. Yu. A holistic lexicon-based approach to opinion mining. In Proceedings of the First ACM International Conference on Web Search and Web Data Mining (WSDM 2008), pages 231–240, 2008.