An efficient web service annotation for domain classification and information retrieval systems using HADLNN classifier. (December 2022)

• Record Type: Journal Article
• Authors: Brindha Merin, J; Aisha Banu, W
• Is Part Of: Advances in engineering software. Volume 174(2022)
• Publication Date: 2022-12
• Subjects: Web Service Description Language (WSDL) -- Hadoop Distributed File System (HDFS) -- Hybrid Artificial Deep Learning Neural Network (HADLNN) -- Cat swarm Optimization (CSO) and Modified K-means (MK-means); Computer-aided engineering -- Periodicals; Engineering -- Computer programs -- Periodicals; Engineering -- Software -- Periodicals; Periodicals; 620.0028553
• DOI: 10.1016/j.advengsoft.2022.103292
• Languages: English
• ISSNs: 0965-9978

Abstract

Highlights: Aimed at exchanging the dissimilar data betwixt distributed applications, web services (WSs) annotations have progressed as a versatile and cost-effectual solution. Information retrieval (IR) assists in establishing the user-essential related information's searching impacts. H owever, rendering quick and efficient IR is a challenging problem. Also, the existent system yields low accuracy, and as well the training time is high. Next, continual WSs utilizing Hadoop Distributed File System (HDFS) is eliminated. Abstract: Aimed at exchanging the dissimilar data between distributed applications, web services (WSs) annotations have progressed as a versatile and cost-effectual solution. Information retrieval (IR)assists in establishing the user-essential related information's searching impacts. However, rendering quick and efficient IR is a challenging problem. Also, the existent system yields slow accuracy, and as well the training time is high. Aimed at overcoming these problems, implemented an effective WS annotation aimed at domain classification and IR systems utilizing the Hybrid Artificial Deep Learning Neural Network (HADLNN). Firstly, the Semantic annotation (SA)stage is executed that comprises text preprocessing, repetitive data removal, feature extraction, and as well Ontology Construction. The text preprocessing offers the partitioning, stop word removal, and as well the stemming procedure aimed at the WSDL dataset. Next, continual WSs utilizing Hadoop Distributed File System (HDFS)is eliminated. After that, the CFC, confidence, support, and as well entropy attributes are taken out;next, the (Web Ontology Language) OWL files as of ontology construction are generated utilizing the protégé tool. After producing the OWL, the owl file is visualized utilizing Eclipse IDE and extracted the values utilizing the reasoner in the protege tool. After WS annotations, the domain is categorized centered on the connecting of WSs utilizing HADLNN. Lastly, the IR procedure is executed on MK-means that groups identical services as of the categorized domain. Preliminary outcomes exhibit that the system proposed offers efficient performance analogized to the existent techniques.