Semantic Similarity Based Data Alignment Using Ontology and Swarm Intelligence Based Annotating Search Results from Web Databases

Abstract

Deep analysis of the SRRs from Web databases (WDB) plays major important role to find exact searching for each user and improve the web pages by analysis results. The number of web pages returned from the WDB is known search result records (SRRs); it consists of the several numbers of pages. These returned
web pages results becomes complex to analysis each data unit since it consists of several number of data unit .To overcome the problem of web analysis in earlier work proposes an annotation methods where the data and text unit nodes are converted in the form of table, but in these work the data and unit nodes are not semantically measured labeled .In order to conquer these problem in this paper propose an ontology based data alignment method to Semantically label each data units in the annotated table for each and every SRRs returned from WDBs. This ontology based data alignment algorithm extract data and text unit nodes features from the SRRs through PSO and then data are aligned based on the semantically labeled named results from ontology. Then perform the swarm
intelligence based ABC algorithm to select best search results from annotated table result through the colony of artificial bees .The colony of the ABC consists of the three major bees such as employed, onlookers and scouts bees. Each and every bee in the ABC selects best annotated search results through the behavior of the ABC for annotated table. Then, for each selected search results annotated it again in different point of the view and combined the dissimilar annotations to anticipate an absolute annotation label intended for it. Our experiments results indicates that the proposed DA-ABC algorithm have high precision and recall result for book application when compare to existing data alignment results.