HBlogRec: A Hybridized Cognitive Knowledge Scheme for Blog Recommendation infusing XGBoosting and Semantic Intelligence
V. Adithya, Gerard Deepak
Abstract
In recent times Blogs have become a mass media and the number of blogs is increasing exponentially. As the number of blogs starts to increase, finding the desired blog based on the user's content is necessary so a recommender system is the need of the hour. Although considerable research has been carried out to provide a reliable recommender system, only a few have proposed approaches that are semantically driven; for this type of multiple domain knowledge problem, semantic intelligence and real-world knowledge enrichment is required. This paper proposes a hybrid architecture that integrates cognitive knowledge addition, XGBosting for recommending blogs, based on semantic similarity comparison for which measures like SemantoSim and K-L divergence are used and the results are optimized under the purview of metaheuristics optimization. The proposed HBlogRec records an Accuracy and Precision of 94.27 % and 96.34% respectively. Which is compared and analyzed with benchmark and baseline approaches and is inferred to be superior in terms of performance.