Implementing Artificial Intelligence in H-BIM Using the J48 Algorithm to Manage Historic Buildings. Issue 8 (13th September 2020)

• Record Type: Journal Article
• Authors: Bienvenido-Huertas, David; Nieto-Julián, Juan Enrique; Moyano, Juan José; Macías-Bernal, Juan Manuel; Castro, José
• Is Part Of: International journal of architectural heritage. Volume 14:Issue 8(2020)
• Page Start: 1148
• Page End: 1160
• Publication Date: 2020-09-13
• Subjects: atificial intelligence -- decision trees -- geometric description language -- J48 algorithm -- H-BIM project -- Pavilion of Charles V; Architecture -- Conservation and restoration -- Periodicals; Historic buildings -- Periodicals; Historic preservation -- Planning -- Periodicals; 721.0288
• DOI: 10.1080/15583058.2019.1589602
• Languages: English
• ISSNs: 1558-3058

Abstract

ABSTRACT: The preservation of the architectural heritage is characterized by the intervention of different technicians, who may disagree on decision-making criteria. In recent years, the H-BIM methodology has emerged to manage these buildings, although the multidisciplinary technical personnel make the decision-taking something of a challenge. In this regard, artificial intelligence may be an opportunity to establish automatic responses, thus optimizing the process. This article proposes a methodology to implement models of classification using the J48 algorithm in a H-BIM model. The case study was focused on a tiles panel from a building which belongs to the Real Alcazar of Seville. First, a model of classification was developed to estimate the degree of intervention with an adequate degree of adjustment. Then, the model was implemented in the H-BIM software by programming using GDL. This methodology automates the decision-making and reduces times of assessments, visualizing and managing the information in the H-BIM model.