Abstract
Context aware systems strive to make more applications intelligent and responsive based on personalised and effective resolutions. It is known that the environment is dynamic and uncertain; therefore, context aware systems require highly dynamic systems to manipulate and function. A genetic algorithm is an evolutionary-based algorithm that operates on a fitness function, crossover and mutation within different populations. An increase in population size reduces the performance of genetic algorithms, leading to bloating, non-convergence and decreased efficiency. A semantic-tagbased genetic algorithm processes the entire population with meaningful tags, which help the algorithm’s functions categorise parents and produce the fittest offspring, thereby improving system performance. Therefore, this paper proposes semantic-tag-based genetic programming for context aware systems, which assigns semantic tags to contexts and applies a genetic algorithm. The proposed work was simulated in the Python environment using the Banking dataset from Kaggle, which showed 91.86% accuracy. Redundant data processing, a lightweight framework, an accurate system with semantic tags, crossovers and mutations in the genetic algorithm for context aware systems is achieved through the proposed work. Performance metrics such as confusion matrix, accuracy, precision, recall and region of convergence are experimented with and the results forecast the improvement in the proposed work in comparison with existing work.
Keywords: Context Aware Systems, Genetic Algorithm, Genetic Programming, Innovation, Optimization, Semantic Tags.