There arises a need for semantically driven framework for online blog tagging which is compliant with Web 3.0 which is currently the semantic standard of the World-Wide Web. The current paper proposes a strategic framework focusing on blog tag generation that encompasses the TF-IDF model onto the blog dataset to extract the informative terms, which is further enriched using CYC, Wikidata and NELL knowledge stores as repositories. The classification model encompassed is the logistic relation classifier, which inputs the enrichments from the CYC, Wikidata and NELL pipeline. The classify instances of the dataset is further used in the model, the category from the dataset is also subjective to generation of metadata, which is further classified using the GAN’s, so we get more permeable and atomic into the framework, which further helps in driving a semantic network. Semantic similarities computed using SimRank under differential evolution. Differential evolution and Morisita’s overlap index, SimRank and Jaccard similarity measures at different stages and levels which yields semantic similarity-based reasoning to yield the best-in-class tags compared to other baseline models. The proposed framework achieves the precision of 95.74% with an F-measure of 95.91% with the least value of FDR.