PICO Extraction by combining the robustness of machine-learning methods with the rule-based methods

被引:0
作者
Chabou, S. [1 ]
Iglewski, M. [1 ]
机构
[1] Univ Quebec Outaouais, Dept Comp Sci & Engn, Gatineau, PQ J8Y 3G5, Canada
来源
2015 World Congress on Information Technology and Computer Applications (WCITCA) | 2015年
关键词
machine-learning methods; rule-based methods; PICO extraction; CRFs;
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
machine-learning methods (MLMs) are robust methods in the extraction of the information; they have been also used in the extraction of PICO elements in order to answer clinical questions; MLMs are only used at coarse-grained level in PICO extraction, because of lack of training corpora for PICO at the fine-grained level. Coarse-grained level cannot explore the semantics within the sentence for use as a means of relevance between different answers. We propose a hybrid approach combining the robustness of MLMs and the fine grained level of RBMs to enhance PICO extraction process and facilitate the validity and the pertinence of the answers to clinic questions formulated with the PICO framework.
引用
收藏
页数:4
相关论文
共 13 条
[1]  
Amini Iman, 2012, OVERVIEW ALTA 2012 S
[2]  
[Anonymous], 2013, CRF YET AN CRF TOOLK
[3]   Combining classifiers for robust PICO element detection [J].
Boudin, Florian ;
Nie, Jian-Yun ;
Bartlett, Joan C. ;
Grad, Roland ;
Pluye, Pierre ;
Dawes, Martin .
BMC MEDICAL INFORMATICS AND DECISION MAKING, 2010, 10
[4]  
Boudin Florian, 2010, P ECIR 2010 C, P50
[5]  
Boudin Florian, 2010, P HUMAN LANGUAGE TEC, V10, P822
[6]  
CHUNG GY, 2009, SENTENCE RETRIEVAL A, V9, P10, DOI DOI 10.1186/1472-6947-9-10
[7]  
Dawes Martin, 2007, Inform Prim Care, V15, P9
[8]  
Demner-Fushman Dina, 2007, ANSWERING CLIN QUEST, V310, P1
[9]  
Hansen M., 2008, P TROMS TEL EHEALTH
[10]  
Huang Xiaoli, 2006, AMIA 2006 S P, V359