An effective multilingual retrieval with query optimization using deep learning technique

Abstract

• Multilingual retrieval is the information that are retrieved in multiple languages, for the response to the query given by the user. • In a cross lingual search, the user query is given to the various information sources thus enabling the retrieved results of different languages. • The single retrieved results would term to be the results merging problem of multilingual search aspects. • The presented model involves processes namely multilingual search and query optimization. • The stacked autoencoder (SAE) model is used for Multilingual Search process. Multilingual retrieval is the information that are retrieved in multiple languages, for the response to the query given by the user. In a cross lingual search, the user query is given to the various information sources thus enabling the retrieved results of different languages. The single retrieved results would term to be the results merging problem of multilingual search aspects. This paper presents a novel query optimization enabling the user to provide a query in Tamil language and the results are retrieved for across languages. The presented model involves processes namely multilingual search and query optimization. The stacked autoencoder (SAE) model is used for Multilingual Search process. In addition, the query optimization process considered as a NP-hard problem, can be resolved by the use of gray wolf optimization (GWO) algorithm. Besides, global vectors (GloVe) technique is applied to construct the domain-specific sentiment lexicon. An extensive set of experimentation analysis was performed and the results are investigated under distinct aspects. The resultant experimental value of the proposed technique in multilingual search obtains the accuracy of 75% and the optimization algorithm gains the supremacy of the existing algorithms.