CAO Jun-Kuo1,2, SHEN Chao1, HUANG Xuan-Jing1, WU Li-De1
1.Department of Computer Science and Engineering, Fudan University, Shanghai 200433 2.School of Information Engineering, Nanchang Hangkong University, Nanchang 330063
Abstract Margin infused relaxed algorithm (MIRA) is an improved ultraconservative algorithm, which is successfully used in classification, ranking and regression. The k-best MIRA (K-MIRA) and dynamic k-best MIRA (DK-MIRA) are proposed. The improved MIRA reduces the optimization constraints progressively as training moves forward. The experiment is carried out on the task of sentence ranking in definitional question answering with K-MIRA and DK-MIRA. The experimental results show that the proposed algorithms greatly improve the performance.
[1] Black P E. Online Algorithm // Black P E, ed. Dictionary Algorithms and Theory of Computation Handbook. Boca Raton, USA: CRC Press LLC, 1999: 10-17, 19-27 [2] Zhao Yu, Xi Hongsheng, Wang Zilei, et al. Online Training of SVM for Multiuser Detection. Journal of System Simulation, 2006, 18(1): 50-53 (in Chinese) (赵 宇,奚宏生,王子磊,等.基于在线SVM的多用户检测算法及仿真.系统仿真学报, 2006, 18(1): 50-53) [3] Dai Liuling, Huang Heyan, Chen Zhaoxiong. An Online Learning Algorithm for Text Categorization with Support Vector Machines. Journal of Chinese Information Processing, 2005, 19(5): 11-16 (in Chinese) (代六玲,黄河燕,陈肇雄.一种文本分类的在线SVM学习算法.中文信息学报, 2005, 19(5): 11-16) [4] Gu Ping, Zhu Qingsheng. A Parallelizable Bayesian Ensemble Online Learning Algorithm. Computer Science, 2006, 33(4): 159-161 (in Chinese) (古 平,朱庆生.一种可并行的贝叶斯集合在线学习算法.计算机科学, 2006, 33(4): 159-161) [5] Crammer K, Singer Y. Ultraconservative Online Algorithms for Multiclass Problems. Journal of Machine Learning Research, 2003, 3(5): 951-991 [6] Crammer K, Singer Y. Online Passive-Aggressive Algorithms. Journal of Machine Learning Research, 2006, 7(4): 551-585 [7] Lin D. MINIPAR: A Minimalist Parser [EB/OL]. [1999-03-12]. http://129.2.53.113.events/colloquia/archives/Spring99/lin.html [8] McDonald R. Discriminative Sentence Compression with Soft Syntactic Evidence // Proc of the Workshop on Cross-Language Knowledge Induction. Trento, Italy, 2006: 297-304 [9] McDonald R, Crammer K, Pereira F. Online Large-Margin Training of Dependency Parsers // Proc of the 43rd Annual Meeting on Association for Computational Linguistics. Ann Arbor, USA, 2005: 91-98 [10] Taskar B, Guestrin C, Koller D. Max-Margin Markov Networks // Proc of the 17th Annual Conference on Neural Information Processing Systems. Vancouver, Canada, 2003: 25-32 [11] Taskar B, Klein D, Collins M, et al. Max-Margin Parsing [EB/OL]. [2004-05-20]. http://acl.ldc.upenn.edu/acl2004/emnlp/pdf/Taskar.pdf [12] McDonald R, Wells K H T N M, Reynar J. Structured Models for Fine-to-Coarse Sentiment Analysis // Proc of the 45th Annual Meeting of the Association of Computational Linguistics. Prague, Czech Republic, 2007: 432-439 [13] Xu Jun, Cao Yunbo, Li Hang, et al. Ranking Definitions with Supervised Learning Methods // Proc of the 14th Conference on International World Wide Web. Chiba, Japan, 2005: 811-819 [14] Han K S, Song Y I, Rim H C. Probabilistic Model for Definitional Question Answering // Proc of the 29th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. Seattle, USA, 2006: 212-219 [15] Voorhees E M. Overview of the TREC 2003 Question Answering Track // Proc of the 12th Text Retrieval Conference. Monterey, USA, 2003: 54-68 [16] Voorhees E M. Overview of the TREC 2004 Question Answering Track [EB/OL]. [2004-10-21]. http://trec.nist.gov/pubs/trec13/papers/QA.OVERVIEW.pdf [17] Kor K W, Chua T S. Interesting Nuggets and Their Impact on Definitional Question Answering // Proc of the 30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. Amsterdam, Netherlands, 2007: 335-342 [18] Cui Hang, Kan M Y, Chua T S. Generic Soft Pattern Models for Definitional Question Answering // Proc of the 28th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. Salvador, Brazil, 2005: 384-391
2009, Vol. 22 
