隨著網路的快速成長，資料探勘(Data Mining)及文字探勘(Text Mining)所須分析的資料集越來越龐大，透過單一機器執行資料探勘分析受限於記憶體大小及其計算能力，不僅運算時間大幅增加，分析資料集的檔案大小也因而受到限制；語意註解萃取出文件的重要內容，凸顯主題加強資料探勘及文字探勘的效果，而資料探勘、文字探勘和語意註解背後都牽涉到大規模的資料處理，透過雲端運算的技術使負載平衡，將運算工作分散至運算叢集中的每一台電腦，不僅加快運算和儲存的速度，更可降低整體的風險。
本研究使用Hadoop軟體實作雲端文字探勘平台，用於分散式文字探勘及結果分析，採用涵蓋21578篇新聞文件的路透社資料集(Reuters 21578)進行實證分析，依照Mod Apte切分法分為訓練資料集及測試資料集用以進行文件分類，文件分類的步驟分為數個部分，分別為進行資料格式轉換的資料前置處理、針對文件內容加註更詳盡的連結及描述的語意註解、用以產生分類預測模型的分類器(簡單貝氏分類器、餘集簡單貝氏分類器)與評估文件分類結果的評估器；路透社資料集經過去除停用字、附加語意註解資料及文本詞彙長度統計分類，再進行簡單貝氏分類器及餘集簡單貝氏分類器的訓練，比較測試資料集的分類正確率作為文件分類實證結果。
本研究根據實驗結果發現，探討去除停用字、語意註解、文件分類演算法及文本詞彙長度對於文件分類正確率的影響：(1)去除停用字使出現頻率高的停用字對於分類預測產生負面影響；(2)語意註解作為詮釋資料的取得方式，可增加文件分類的效果；(3)餘集簡單貝氏分類器，可用以減少偏斜資料對於分類預測結果的誤判；(4)文本詞彙長度較長的文章則會某種程度主導分類預測結果，造成誤判的產生，降低分類正確率；透過上述各影響因子的調整使文件分類的結果得到改善，使得文件分類正確率獲得較佳的效果。
本研究提出之系統以雲端運算環境運行文件分類演算法，使得大型資料集得以更為迅速取得分析結果，使用語意註解作為詮釋資料的來源，使得文件分類模型產生過程中有更多資訊可分析，使得機器判斷的正確程度獲得改善，亦可將文件轉換為語意網文件，供語意網搜尋引擎查詢檢索，未來應加入Twitter或Facebook等擁有大量非結構化資料的網站之資料，使本平台得以分析更大規模的資料，並且考慮資料集類別分佈的集中程度對分類正確率的影響程度，同時應實作效果更佳的分類演算法，進而改善系統整體的結果。

Nowadays, businesses perform data mining and text mining need to handle large scale dataset. The computational resources of servers are often limited and lack of efficient to compute analytical jobs. But if they could run their data mining jobs under cloud computing clusters, they are able to get results very quickly on a large dataset without "out of memory" problems.
In this paper, a series of experiments are conducted to measure and analyze the accuracy of the classification algorithms implemented on Hadoop using Reuters-21578 dataset; the process of text mining consisted of four stages: (1)data preprocessing, (2)semantic annotation, (3)classifier, (4)evaluator. Reuters-21578 had divided into training set and testing set based on Mod Apte Split, processed by stopwords removal, appended semantic annotations as metadata and splitted into several subsets according to different document sizes. Experiments outlined several issues that will need to be considered when conducting text mining.
According to the experiment results, the researcher found that stopwords removal, semantic annotation, different classification algorithms and different document sizes could improve the classification accuracy. First, stopwords removal avoids common words from becoming noises that will do harm to classification result. Second, semantic annotation as the extra information could improve the result. Third, complementary naive bayes algorithm could solve the decision boundary problem which naive bayesian cannot handle. Fourth, long documents could dominate the classification results. Sixth, the class imbalance problem could cause a drop of classification accuracy. Text mining result could be improved by adjusting the parameters found above.

[1] Apte, C., Damerau, F., & Weiss, S. M. (1994). Towards language independent automated learning of text categorization models. Paper presented at the ACM SIGIR Conference on Research and Development in Information Retrieval.
[2] AWS in Education. (2010). from http://aws.amazon.com/education
[3] Berendt, B., Hotho, A., & Stumme, G. (2002). Towards semantic web mining. The Semantic Web—ISWC 2002, 264-278.
[4] Center, N. R. (2008). DisCo. from http://discoproject.org/
[5] Chang, F., Dean, J., Ghemawat, S., Hsieh, W. C., Wallach, D. A., Burrows, M., et al. (2008). Bigtable: A distributed storage system for structured data. ACM Transactions on Computer Systems (TOCS), 26(2), 4.
[6] Chu, C. T., Kim, S. K., Lin, Y. A., Yu, Y. Y., Bradski, G., Ng, A. Y., et al. (2007). Map-reduce for machine learning on multicore. Paper presented at the NIPS.
[7] Dean, J., & Ghemawat, S. (2008). MapReduce: Simplified data processing on large clusters. Communications of the ACM, 51(1), 107-113.
[8] Dlugolinsky, S., Laclavik, M., Seleng, M. (2010). Ontea Semantic Annotation. from http://ontea.sourceforge.net/
[9] Fayyad, U., Piatetsky-Shapiro, G., & Smyth, P. (1996). The KDD process for extracting useful knowledge from volumes of data. Communications of the ACM, 39(11), 27-34.
[10] Foundation, A. S. (2008). Cassandra. from http://incubator.apache.org/cassandra/
[11] Foundation, A. S. (2008). Hadoop. from http://hadoop.apache.org/core/
[12] Foundation, A. S. (2010). Hbase. from http://hadoop.apache.org/hbase/
[13] Gillick, D., Faria, A., & DeNero, J. (2006). Mapreduce: Distributed computing for machine learning.
[14] Google. (2010). Google App Engine. from http://code.google.com/intl/en/appengine
[15] Hypertable. (2010). Hypertable. from http://www.hypertable.org
[16] Jenkin, N. (2009). COMP390-09A Report Distributed Machine Learning with Hadoop.
[17] Kibriya, A., Frank, E., Pfahringer, B., & Holmes, G. (2005). Multinomial naive bayes for text categorization revisited. AI 2004: Advances in Artificial Intelligence, 235-252.
[18] Laclavik, M., eleng, M., & Hluchy, L. (2008). Towards large scale semantic annotation built on mapreduce architecture. Computational Science–ICCS 2008, 331-338.
[19] Laclavik, M., Seleng, M., Gatial, E., Balogh, Z., & Hluchy, L. (2007). Ontology based Text Annotation–OnTeA. Information modelling and knowledge bases XVIII, 311.
[20] Lioma, C., Moens, M. F., & Azzopardi, L. (2008). Collaborative annotation for pseudo relevance feedback. ESAIR, 11, 25-35.
[21] Maedche, A. (2001). Ontology learning for the semantic web: Intelligent Systems, IEEE.
[22] Mell, P., & Grance, T. (2009). The nist definition of cloud computing. National Institute of Standards and Technology.
[23] Ontotext. (2009). KIM Semantic Annotation. from http://www.ontotext.com/kim/introduction.html
[24] Papadimitriou, S., & Sun, J. (2008). Disco: Distributed co-clustering with Map-Reduce: A case study towards petabyte-scale end-to-end mining. Paper presented at the ICDM.
[25] ReadWriteWeb. (2010). Does Facebook Really Want a Semantic Web? , from http://www.readwriteweb.com/archives/does_facebook_really_want_a_semantic_web.php
[26] Reeve, L., & Han, H. (2005). Survey of semantic annotation platforms. Paper presented at the Proceedings of the 2005 ACM symposium on Applied computing.
[27] RIGHTSCALE. (2010). RIGHTSCALE. from http://www.rightscale.com/index.php
[28] Sivashanmugam, K., Sheth, A., Miller, J., Verma, K., Aggarwal, R., & Rajasekaran, P. (2003). Metadata and semantics for Web services and processes. Datenbanken und Informationssysteme: Festschrift zum, 60, 245-271.
[29] Stanford. (2005). TAP. from http://ksl.stanford.edu/projects/TAP/
[30] Stanford. (2007). Phoenix. from http://csl.stanford.edu/~christos/sw/phoenix/
[31] Wegener, D., Mock, M., Adranale, D., & Wrobel, S. (2009). Toolkit-based high-performance data mining of large data on MapReduce clusters.
[32] Wikipedia. (2007). Dbpedia. from http://wiki.dbpedia.org/
[33] Wikipedia. (2010). Cloud Computing. from http://en.wikipedia.org/wiki/Cloud_computing
[34] Witten, I. H., & Frank, E. (2005). Data Mining: Practical machine learning tools and techniques: Morgan Kaufmann Pub.
[35] gipi的學習筆記，2009，http://www.dotblogs.com.tw/jimmyyu/。
[36] 劉繼鴻，2009，影音Web 2.0平台網站上行銷傳播之社會網絡與資料探勘分析研究－以YouTube-Mac網絡為例，國立政治大學資訊管理研究所碩士論文。
[37] 劉俊宏，2009，雲端運算環境下學習社群服務導向架構平台之研究，國立政治大學資訊管理研究所碩士論文。
[38] 王耀聰、陳威宇，2008，雲端運算簡介，http://bit.ly/bXsTVT。
[39] 謝良奇，2008，HP、Intel、Yahoo共組開放源碼雲端運算計畫，http://bit.ly/adLC7D。
[40] 葉慶隆，2009，Semantic Web and Knowledge Management， http://www.deg.byu.edu/ding/research/SemanticAnnotation.html。
[41] 陳瀅，2010，雲端策略：雲端運算與虛擬化技術，天下雜誌。