Design of an Automated Construction Platform for Advanced Mathematics Content Integrating Knowledge Graph and Generative Artificial Intelligence

Abstract

This paper proposes an automated construction platform for advanced mathematics content that integrates knowledge graphs and generative artificial intelligence. Based on domain ontologies, it realizes semantic closed modeling of propositions, formulas, and reasoning paths. Combining graph neural networks and symbolic constraints, it completes interpretable and multi-hop reasoning. It uses multi-modal encoding and structure-controlled decoding to generate and verify mathematical content, and constructs a full-link system covering collection, generation, typesetting, and personalized recommendation. The test results show that the platform performs excellently in terms of consistency, verifiability, and engineering controllability, and can support the automatic generation and intelligent application of advanced mathematics knowledge.