Integrating Active Learning Strategy to the Ensemble Kernel-based Method for Protein-Protein Interaction Extraction

LI Lishuang; HUANG Degen; WANG Min; JIANG Zhenchao

Article Contents

Article Navigation > Chinese Journal of Electronics > 2013 > 22(1): 41-45

LI Lishuang, HUANG Degen, WANG Min, et al., “Integrating Active Learning Strategy to the Ensemble Kernel-based Method for Protein-Protein Interaction Extraction,” Chinese Journal of Electronics, vol. 22, no. 1, pp. 41-45, 2013,

Citation:

LI Lishuang, HUANG Degen, WANG Min, et al., “Integrating Active Learning Strategy to the Ensemble Kernel-based Method for Protein-Protein Interaction Extraction,” Chinese Journal of Electronics, vol. 22, no. 1, pp. 41-45, 2013,

Citation:

PDF( 319 KB)

Integrating Active Learning Strategy to the Ensemble Kernel-based Method for Protein-Protein Interaction Extraction

School of Computer Science and Technology, Dalian University of Technology, Dalian 116023, China

Funds: This work is supported by the National Natural Science Foundation of China (No.61173101, No.61173100).

Received Date: 2011-12-01
Rev Recd Date: 2012-02-01
Publish Date: 2013-01-05

Abstract

Abstract

This paper presents an ensemble kernelbased active learning method for PPI (Protein-protein interaction) extraction. This ensemble kernel is composed of feature-based kernel and structure-based kernel. Experimental results show that the F-scores of PPI extraction using ensemble kernel model on AIMED (Abstracts in medline), IEPA (the Interaction extraction performance assessment corpus) and BCPPI (Biocreative PPI dataset) corpora are 64.50%, 69.74% and 60.38% respectively. As the passive learning methods need large labeled data sets and it is expensive to label data manually, we integrate active learning strategy into the ensemble kernel model. The uncertainty-based sampling strategy is used in the active learning method. Two experiments for active learning are conducted on AIMED, IEPA, BCPPI corpus. The experimental results integrating the active learning strategy show that the F-scores on AIMED, IEPA and BCPPI corpora are better than those using the passive learning, and meantime reduce the labeling data.
- Protein-protein interaction (PPI),
- Combined kernel,
- Active learning,
- SVM

FullText(HTML)

References(13)

References

K. Fundel, R. Küffner, R. Zimmer, “RelEx-Relation extractionusing dependency parse trees Place”, Bioinformatics, Vol.23,No.3, pp.365, 2007.

R. Bunescu, R. Mooney, A. Ramani, E. Marcotte, “Integratingco-occurrence statistics with information extraction for robustretrieval of protein interactions from Medline Place”, Proc. ofBioNLP-2006, New York City, America, pp.49-56, 2006.

B. Liu, et al., “Dependency-driven feature-based learning forextracting protein-protein interactions from biomedical textPlace”, Proc. of COLING’10 Proceeding of the 23rd InternationConference on Computational Linguistics, Beijing, China,pp.757-765, 2010.

Z.H. Yang, et al., “Multiple kernel learning in protein-proteininteraction extraction from biomedical literature Place”, ArtificialIntelligence in Medicine, Vol.51, No.3, pp.163-173, 2011.

L.S. Li, et al., “A hybrid model combining CRF with boundarytemplates for Chinese person name recognition Place”, InternationalJ. Advanced Intelligence, Vol.2, No.1, pp.73-80, 2010.

L.S. Li, D.G. Huang, D. Li, “Recognizing Chinese person namesbased on hybrid models Place”, International Journal, Vol.3,No.2, pp.219-228, 2011.

Z.R. Zhou, X.F. Song, M.H. Wang, “Predicting protein-proteininteractions based on ensemble classifiers Place”, Acta ElectronicaSinica, Vol.38, No.6, pp.1464-1467, 2010.

X. Zhang, et al., “Extracting protein-protein interaction frombiomedical literature using an ensemble kernel”, Journal of Informationand Computational Science, Vol.6, No.2, pp.1067-1075, 2009.

L.S. Li, J.Y. Ping, D.G. Huang, “Protein-protein interactionsextraction from biomedical literatures”, Journal of Informationand Computational Science, Vol.7, No.5, pp.1065-1073, 2010.

R. Bunescu, et al., “Comparative experiments on learning informationextractors for proteins and their interactions Place”,Artificial Intelligence in Medicine, Vol.33, No.2, pp.139-155,2005.

I. Xenarios, et al., “DIP: The database of interacting proteins:2001 update Place”, Nucleic Acids Research, Vol.29,No.1, pp.239-241, 2001.

B.J. Cui, H.F. Lin, Z.H. Yang, “Uncertainty sampling-basedactive learning for protein-protein interaction extraction frombiomedical literature Place”, Expert Systems with Applications,Vol.36, No.7, pp.10344-10350, 2009.

Y.L. He, K. Nakata, D.Y. Zhou, “Ontology-based proteinproteininteractions extraction from literature using the hiddenvector state model Place”, Proc. of Data Mining Workshops,2008. ICDMW’08IEEE, Pisa, Italy, pp.736-743, 2008.

Relative Articles

Supplements(0)

Cited By

Proportional views