The GraphAware Natural Language Processing (NLP) plugin, developed as part of the GraphAware Enterprise Reco plugin for Neo4j, is used to deliver high-quality recommendations to end users by combining content-based and ontology-based cosine similarities with a collaborative filtering approach. The plugin models documents using three dimensions: Content, Tag, and User, providing different views of the document's meaning. It uses a combined approach to compute document similarity, creating three vectors for each document based on its internal content, tags, and user interactions. These vectors are then used to calculate a combined similarity score, which represents new knowledge extracted from the data available in the graph database. The plugin is open-sourced under GPL and aims to provide a complete end-to-end customized search framework with other plugins available on the GraphAware products page.