Building a Neo4j graph database for building materials recommendations in the process of building renovation

Abstract

Modern building renovation processes require optimizing the choice of building materials, taking into account many interrelated factors: energy efficiency, environmental friendliness, compliance with technical requirements and economic feasibility. Traditional relational databases are limited in modeling complex relationships between building materials, their characteristics, and contextual application conditions. This study presents an innovative approach to the development of a building materials recommendation system based on the Neo4j graph database, which takes into account the multidimensional nature of the relationships between the elements of the construction industry. The work uses a comprehensive methodology, including the analysis of the requirements for materials during renovation, the design of graph model ontology, the development of recommendation algorithms and experimental validation on real renovation projects of commercial and residential buildings. The results demonstrate a 37.8% increase in the accuracy of recommendations compared to traditional methods, a 42.3% reduction in decision-making time, and a 28.6% improvement in the compliance of selected materials with technical requirements. The practical application of the developed solution at construction sites showed a reduction in the cost of purchasing materials by 18.4% and an increase in energy efficiency of renovated buildings by 21.3%. The proposed system opens up new opportunities for the digitalization of renovation processes and can be integrated with existing BIM platforms and construction management systems.