[Background.] The natural language processing for requirements engineering (NLP4RE) community introduced NLP4RE ID Cards to describe empirical research practice from scientific publications and foster replicability. However, these artifacts are currently stored as static PDF documents, which limits accessibility, impedes machine actionability, and prevents efficient reuse, e.g., for joint analyses. [Aims.] We aim to transform these static documents into a findable, accessible, interoperable, and reusable (FAIR) representation format using a research knowledge graph (RKG). In particular, we organize the NLP4RE ID Cards in the Open Research Knowledge Graph (ORKG) to improve access to the knowledge and its reuse. [Method.] We develop a pipeline to extract, normalize, and structure data from NLP4RE ID Cards. This data is ingested into the ORKG using a tailored semantic graph schema. Furthermore, we integrate the data in the neuro-symbolic dashboard EmpiRE-Compass that combines symbolic SPARQL querying with neural large language models (LLMs) to facilitate knowledge exploration, synthesis, and reuse. [Results.] With the pipeline, we successfully migrated 50 NLP4RE ID Cards into a persistent, retrievable dataset in the ORKG. The dashboard integration allows researchers to answer complex competency questions — such as identifying common evaluation metrics or dataset properties across all NLP4RE ID Cards — dynamically. The extraction process also reveals significant limitations in the original PDF-based workflow regarding data consistency. [Conclusions.] Transitioning to an open science infrastructure, such as the ORKG, significantly enhances the utility of the scientific knowledge that was encapsulated in the PDF documents. To ensure long-term sustainability, we propose extending the static NLP4RE ID Cards in PDF format with direct, schema-driven web interface and LLM-assisted data entry via EmpiRE-Compass and the ORKG.