LDAC 2023: Advancing Linked Building Data

Why LDAC 

The Linked Data in Architecture and Construction (LDAC) 2023 conference was a nexus of innovation, collaboration, and cutting-edge research, captivating the minds of professionals and enthusiasts in architecture and linked building data domains. Within this vibrant environment, ONTOText, Birmingham City University and BuildingSmartInternational joined a collective of experts and visionaries and played a pivotal role by presenting groundbreaking papers and contributing to the conference’s vibrant discourse. 

ACCORD’s Active Participation 

The ACCORD project vision is strongly based on the linked building data exploitation for data models and rules expression. Being deeply involved in the domain development, consortium stood at the forefront of LDAC 2023, showcasing remarkable contributions highlighting their commitment to shaping the future of linked building data. Anchored by a shared vision of advancing the architecture, engineering, construction, and operations (AECO) sector through semantic technologies, ACCORD’s involvement left an indelible mark on the event. 

Summer School 

At the heart of ACCORD’s engagement, Vladimir Alexiev from Ontotext, an esteemed figure within the consortium, demonstrated his unwavering commitment by supporting the LDAC Summer School. Dedication to fostering the growth of young talent and nurturing the next generation of professionals underscored the consortium’s collaboration and knowledge dissemination ethos.  

Read more: https://linkedbuildingdata.net/ldac2023/summerschool/  

Unveiling the Research Papers 

ACCORD’s presence was amplified by presenting three pioneering research papers. Each addresses critical aspects of linked building data and its transformative potential within the AECO sector.  

Semantisation of Rules for Automated Compliance Checking  

Authors: Edlira Vakaj, Maxime Lefrançois, Amna Dridi, Thomas Beach, Mohamed Gaber, Gonçal Costa Jutglar, He Tan  

This paper delved into the intricate realm of automated compliance checking, where rules were semantified to facilitate efficient assessment. The authors’ work illuminated pathways for enhancing compliance processes through semantic technologies and is promising to streamline and elevate the industry’s quality assurance practices.  

Abstract  

Presentation (PDF)  

Semantic bSDD: Improving the GraphQL, JSON, and RDF Representations of buildingSmart Data Dictionary  

Authors: Vladimir Alexiev, Mihail Radkov, Nataliya Keberle  

A transformative effort to enhance the representation of the buildingSmart Data Dictionary (bSDD). By optimising GraphQL, JSON, and RDF representations, the paper aimed to boost data interoperability, ultimately fortifying the AECO domain’s ability to harness valuable insights.  

Abstract  

Full paper (PDF)  

Presentation (PDF)    

Modular Knowledge Integration for Smart Building Digital Twins – awarded paper 

Authors: Isaac Fatokun, Arun Raveendran Nair, Thamer Mecharnia, Maxime Lefrançois, Victor Charpenay, Fabien Badeig, Antoine Zimmermann  

This collaborative paper addressed the intricate challenge of integrating modular knowledge within smart building digital twins. By leveraging interconnected knowledge, the authors aimed to enhance the accuracy and efficiency of these digital replicas, underscoring their crucial role in modern architectural processes.  

Abstract  

Full paper (PDF)  

Presentation (PDF)  

 

Leveraging Word Embeddings and Transformers to Extract Semantics from Building Regulations Text 

Authors: Odinakachukwu Okonkwo, Amna Dridi and Edlira Vakaj  

This paper explores using advanced Natural Language Processing (NLP) techniques to extract meaningful insights from unstructured text in the Architecture, Engineering, and Construction (AEC) domain. The study focuses on British building regulations, demonstrating how NLP methods like word embeddings and transformers, including word2vec, BERT, and SBERT, can accurately capture semantic patterns. The results reveal high accuracy levels, both at the word (80%) and sentence (100%) levels, indicating the potential of modern NLP for understanding complex regulatory texts. 

Abstract 

Full paper (PDF) 

Presentation (PDF)

An Inspiring Commitment 

ACCORD is committed to advancing the AECO domain through a harmonious blend of expertise and collaboration. The consortium’s presence at LDAC 2023 was a testament to their dedication to fostering innovative solutions, nurturing emerging talent, and transforming the industry’s landscape. As a result, ACCORD’s vision could be validated. It will continue to shape the future of linked building data, propelling the AECO sector to new heights of excellence.