（1.國防科學技術大學信息系統工程重點實驗室，湖南長沙 410073；2.中國航天員科研訓練中心，北京 100094）
摘要：本文將“辯證推理”引入到數據挖掘所發現的規則的自動知識評估中，通過將多Agent的經驗數據相分離，各Agent利用各自的經驗數據基于擂臺辯論模型展開辯論，實現對數據挖掘知識的自動評估。本方法采用了數據挖掘得到的關聯規則作為論據，并使用了一種新型的擂臺辯論模型。本文基于辯證推理實現多方論據博弈，探索解決數據挖掘知識的自動分析評估與篩選的新方法，并通過在分類問題中的應用，在一定程度上提高了知識評估結果的準確性以及評估的效率。
關鍵詞：數據挖掘；關聯規則；辯論；知識評估
中圖分類號：TP18 文獻標識碼：A 文章編號：
Automatic Knowledge Assessment Methodology based on Arena Argumentation
XU Jun-yi1,YAO Li1, LI Le1, LI Peng-jie2
(1. Science and Technology on Information Systems Engineering Laboratory,National University of Defense Technology,Changsha 410073,China. 2. Astronaut Research and Training Center of China, Beijing 100094, China. Corresponding author: XU Jun-yi, E-mail: xujunyi0923@163.com)
Abstract: This article will introduce the "dialectical reasoning" into the automatic knowledge assessment of rules from data mining. By the separation of the empirical data sets of multi-Agent, each Agent use their own experience to argue in the arena model, to achieve the automatic evaluation of the knowledge of data mining. This method used association rules from data mining as an argument and a new model of the arena argumentation. Based on dialectical reasoning to realize multi-party argumentation, this article explore a new method of automatic assessment and screening of knowledge of data mining, and through the application of the classification problem, to some extent, improve the accuracy of the results of the assessment of knowledge and the efficiency of assessment.
Key words: Data mining；Association rules；Argumentation；Knowledge assessment
參考文獻 (References)
[1] G. Piatetsky-Shapiro. Knowledge discovery in databases: 10 years after. In ACM SIGKDD Explorations, vol.1 (2), (2000). pp: 59 – 61.
[2] J. Han and M. Kamber. Data mining: Concepts and techniques (2nd Edition). Morgan Kaufmann, (2006).
[3] Ludwig Jeremy, Livingston Gry. What's New? Using Prior Models as a measure of Novelty in Knowledge Discovery [C]. The Twelfth IEEE International Conference on Tool with Artificial Intelligence, Canada, 2000. 322- 330.
[4] M. Halkidi. Quality assessment and uncertainty handling in data mining process . Proc, EDBT Conference, Konstanz, Germany. 2000.
[5] M. Wardeh, T. Bench-Capon and F.P. Coenen . PADUA: a protocol for argumentation dialogue using association rules. In AI and Law. Springer, vol. 17(3), (2009). pp: 183 – 215.
[6] M. Wardeh, T. Bench-Capon and F.P. Coenen . Multi-Party Argument from Experience. To appear in Proc. 6th Int. Workshop on Argumentation in Multiagent Systems (ArgMAS’09). Budapest, Hungary (2009).
[7] Bench-Capon T. J. M. and Dunne P. E. Argumentation in artificial intelligence. Artificial Intelligence 171 (1), 2007 114-159.
[8] F. Coenen, P. H. Leng, and S. Ahmed. Data structure for association rule mining: T-trees and p-trees. IEEE Trans. Knowl. Data Eng., 16(6):774–778, 2004.
作者簡介：
許珺怡，女，1988年9月出生，安徽合肥人，碩士在讀�，F就讀于國防科學技術大學，專業為管理科學與工程，方向為信息資源管理與智能決策技術。本人主要研究方向為：人工智能，數據挖掘以及辯論理論。
