Abstract:
The rapid growth of online platforms and the abundance of available information have led to an overwhelming number of choices for users. Recommender systems have emerged as vital tools to assist users in finding relevant and personalized recommendations from a vast pool of options. In multi-criteria recommender systems, considering multiple dimensions or criteria, such as user preferences, item characteristics, and contextual factors, is crucial to enhance recommendation accuracy.
This paper proposes a novel approach to enhance the prediction accuracy of a multi-criteria recommender system by incorporating an adaptive genetic algorithm. Genetic algorithms are widely recognized for their ability to optimize complex problems by mimicking the process of natural selection. The proposed adaptive genetic algorithm dynamically adjusts its parameters based on the characteristics of the recommendation dataset, thereby improving the efficiency and effectiveness of the recommendation process.
The key steps of the proposed approach are as follows: (1) preprocessing and feature extraction from the recommendation dataset; (2) encoding the extracted features into a genetic representation; (3) generating an initial population of potential solutions using the genetic encoding; (4) evaluating the fitness of each individual solution based on prediction accuracy; (5) iteratively applying genetic operators, such as selection, crossover, and mutation, to evolve the population and produce new generations of solutions; (6) dynamically adjusting genetic algorithm parameters, such as population size, crossover and mutation rates, based on the recommendation dataset characteristics; (7) terminating the algorithm when a stopping criterion, such as a maximum number of generations or convergence, is met; (8) selecting the best solution as the recommendation output.
Experimental evaluations conducted on a real-world recommendation dataset demonstrate the effectiveness of the proposed approach in enhancing prediction accuracy. The adaptive genetic algorithm consistently outperforms traditional recommendation techniques, such as collaborative filtering and content-based filtering, in terms of accuracy, coverage, and diversity of recommendations.
The outcomes of this research contribute to the advancement of multi-criteria recommender systems by introducing an adaptive genetic algorithm that intelligently adapts to the characteristics of the recommendation dataset. The proposed approach has the potential to benefit various domains, including e-commerce, social media, and personalized content delivery, by providing users with more accurate and diverse recommendations tailored to their preferences and needs.
ENHANCING PREDICTION ACCURACY OF A MULTI-CRITERIA RECOMMENDER SYSTEM USING ADAPTIVE GENETIC ALGORITHM. GET MOREÂ Â COMPUTER SCIENCE PROJECT TOPICS AND MATERIALS
