Pattern Matching Method for Deep Web Interface Integration

CHEN Li-jun,LIN Huai-zhong
DOI: https://doi.org/10.3969/j.issn.1000-3428.2012.12.012
2012-01-01
Abstract:By anglicizing the limitations of existing evidence theory method for Deep Web interface integration,a Deep Web pattern matching method based on concept word and semantic heterogeneity model is proposed.The method preprocesses pattern through extracting concept word,discriminates and combines group attributes to convert m︰n complex matching into 1︰1 simple matching for improving implement efficiency.By introducing instance into semantic heterogeneity model,the problem of mining semantic heterogeneity synonymy attributes is resolved by computing,synthetic evaluating,and selecting similarity values of attribute features.Experimental results indicate that compared with evidence theory method,the efficiency and accuracy of the method is improved obviously.
What problem does this paper attempt to address?