以概念為維度之向量空間模型為基礎以進行文件分群之研究

Document clustering based on vector space model with concepts as the dimension value

蘇千傑(Chien-chieh Su)

指導教授 ： 周世傑

國立中央大學/管理學院/資訊管理學系/碩士(2007年)

摘要

在資訊檢索相關研究中，文件分群是用來令使用者能夠更加快速找到自己所需資訊的技術，利用分群的結構，我們可以有效的管理各種知識與資訊，它是一門知識管理的工具。
文件分群通常需要進行文件相似度比對，傳統上利用文章中的字彙當作向量空間模型的維度，此種方式，有一項弱點，即當兩篇文章在語意上相同，但用不同的字彙呈現時，會無法準確判斷文章間相似度而使文件分群困難。本研究結合了概念擷取與向量空間模式(Vector space model)兩種技術來協助文件分群，希望能夠以文章中所涵蓋的概念來代表文章，然後產生一個以概念為維度的向量空間模型，已進行文件相似度比對，希望能提高文件相似度比對的效能，進而使分群的效果更加完善。
我們進行了實驗來觀察使用概念為維度的向量空間模型，是否比傳統使用字彙為維度的向量空間模型，對於文件分群，具有更佳的效能，結果顯示使用概念為維度的向量空間模型，確實能夠幫助我們對文件作更準確的分群。

關鍵字

知識管理 ； 概念擷取 ； 向量空間模型 ； 文件分群 ； 資訊檢索

並列摘要

In Information Retrieval, document clustering is a technology that can enhance the efficiency in the retrieving of needed information. With document clustering, one can efficiently management all kinds of knowledge and information. Document clustering is a tool for knowledge management.
Traditionally, document clustering is based on document similarity comparison where the document is represented by the vector space model with term as the dimension value. In this approach, the documents with the same semantic meaning might be classified as unsimilar because they are described with different words.In this research, we have integrated the technology of concept extraction with vector space model for document similarity comparison. We extract concepts from the documents first, then create a vector space model with the extracted concepts as the dimension value for the document. Documents similarity comparison is based on the concept-dimensioned vector space model. We wish that the concept based vector space model could enhance the document clustering efficiency.
We have experimented with the document clustering effect for the concept based vector space modle. The results show that the concept based vector space model can perform better than term based vector space model.

並列關鍵字

knowledge management ； concept extraction ； vector space model ； document clustering ； information retrieval

參考文獻

[4] Hung Chim. A new suffix tree similarity measure for document clustering. In Proceedings of the 16th international conference on World Wide Web, pages 121-130, 2007.

[5] N. Fuhr. Probabilistic model in information retrieval. The Computer Journal, 35(3):243-255, 1992.

[7] A. Griffith, H. C. Luckhurst, and P. Willet. Using inter-document similarity information in document retrieval systems. Journal of the American Society for Information Science, 37(1):3-11, 1986.

[10] Nyeint Nyeint Myat and Khin Haymar Saw Hla. A combined approach of formal concept analysis and text mining for concept based document clustering. In Proceedings of the 2005 IEEE/WIC/ACM international conference on web intelligence, pages 330-333, 2005.

[14] W. M. Shaw Jr., R. Burgin, and P. Howell. Performance standards and evaluations in IR text collections: Cluster-based retrieval models. Information processing & management, 33(1):15-36, 1997.