A strategic framework for book recommendations is required for Web 3.0, one that makes use of the web’s actual contents to suggest books based on them. This work therefore provides a data knowledge-centric model that aggregates knowledge from semantic cookies like Wikidata and Google’s book API and populates the title. The model generates ontologies and then further images using CYC. It also outgrows the data set through its contextualization using LDA. The framework primarily focuses on enhancing the query words and matching instances through the Jiang-Conrath similarity, Hulbert index, and APMI adaptive point-wise machine information measure at various stages. Recurrent neural networks are given a robust learning infrastructure to classify the data sets. An overall precision of 93.82% with the F-measure of 94.44% and the lowest FDR of 0.07 has been gained by the suggested model, which makes it the best in-class model for recommendation of books, and which is in adherence to Web 3.0.