A Lite-SVM Based Semantic Search Model for Bigdata Analytics in Smart Cities

The distributed file system used by conventional concepts of big data lacks the extra security and privacy features, which are required more time and energy wastage. In this research, propose a methodology for managing private and safe big data modeling with LITE SVM-based semantic application. It's not possible to scale the big data platform, and there's no assurance of security and time scaling. The suggested technique uses an SVM machine to accurately address these concerns without compromising either security or privacy and without affecting the energy consumption of the application. The main objective of this research is, to manage the Bigdata platform with effective storage in less Time of Conversion (ToC). The existing models like Genetic Algorithm (GA), Differential Annealing (DE) for global optimization, and Fuzzy models were limited to functioning. So that to crossover the above limitations proposed Lite SVM machine learning techniques. The accuracy 82.23%, true positive rate 98.23%, recall 84.53%, & F1 score 92.34% were observed with suggested Lite SVM and compared with past techniques. Such new and enhanced SVM method outcomes are competitive with the most recent methods.