Title :
Ranking SVM for multiple kernels output combination in protein-protein interaction extraction from biomedical literature
Author :
Yang, Zhihao ; Lin, Yuan ; Wu, Jiajin ; Tang, Nan ; Lin, Hongfei ; Li, Yanpeng
Author_Institution :
Dept. of Comput. Sci. & Eng., Dalian Univ. of Technol., Dalian, China
Abstract :
Knowledge about protein-protein interactions unveils the molecular mechanisms of biological processes. This paper presents a multiple kernels learning-based approach to automatically extracting protein-protein interactions from biomedical literature. Experimental evaluations show that our approach can achieve state-of-the-art performance with respect to comparable evaluations, with 64.88% F-score and 88.02% area under the receiver operating characteristics curve (AUC) on the AImed corpus.
Keywords :
biology computing; learning (artificial intelligence); proteins; sensitivity analysis; support vector machines; AImed corpus; F-score; biomedical literature; learning; molecular mechanisms; multiple kernels output combination; protein-protein interaction extraction; ranking SVM; receiver operating characteristics curve; Data mining; Feature extraction; Kernel; Protein engineering; Proteins; Support vector machines; Training; Multiple kernels learning; Protein-protein interaction; Support Vector Machines;
Conference_Titel :
Bioinformatics and Biomedicine (BIBM), 2010 IEEE International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-8306-8
Electronic_ISBN :
978-1-4244-8307-5
DOI :
10.1109/BIBM.2010.5706635