Leveraging Social Media Data to Understand COVID-19 Prevention Measures in Construction: A Machine Learning Approach
The COVID-19 pandemic was a particularly challenging time for the construction industry as it experienced significant disruptions to operations, affecting various stakeholders. With various national and international health agencies promoting preventive measures, the construction industry struggled...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Buildings |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-5309/15/13/2191 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850118784799473664 |
|---|---|
| author | Emmanuel B. Boateng Daniel Oteng Dan N. O. Bonsu Vinod Gopaldasani |
| author_facet | Emmanuel B. Boateng Daniel Oteng Dan N. O. Bonsu Vinod Gopaldasani |
| author_sort | Emmanuel B. Boateng |
| collection | DOAJ |
| description | The COVID-19 pandemic was a particularly challenging time for the construction industry as it experienced significant disruptions to operations, affecting various stakeholders. With various national and international health agencies promoting preventive measures, the construction industry struggled with the implementation of these measures due to the unique nature of the work involved in construction. This study aimed to highlight the ways in which stakeholders in the construction industry interacted and responded to the prescribed preventive measures through social media analysis. Using model-based clustering and structural topic modelling, this study provided insights into the prevalent discussion topics in social media around prevention measures in construction. In addition, sentiment analysis demonstrated interesting polarisation around the topic areas. Four prevalent topics that encapsulated the entirety of the social media data were identified, with two of the topics showing an upward trend, as expected, while the other two topics showed a contrasting downward trend. These findings offer practical value for construction managers and policymakers by revealing the effectiveness of different communication strategies and identifying areas where prevention measures faced resistance or acceptance. The sentiment polarisation patterns (50% positive, 40% negative) provide actionable insights for developing more targeted engagement approaches, while the topic evolution trends inform the timing and focus of safety communications. Construction organisations can leverage these insights to improve workplace safety protocols and enhance stakeholder buy-in for future health initiatives. This study lays the foundation for future studies to investigate the connections between the prevalent prevention and the interrelated dynamics within the conversation regarding COVID-19 prevention strategies in the construction sector. |
| format | Article |
| id | doaj-art-e74cff5228624b42afbefe7661dc8c1c |
| institution | OA Journals |
| issn | 2075-5309 |
| language | English |
| publishDate | 2025-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Buildings |
| spelling | doaj-art-e74cff5228624b42afbefe7661dc8c1c2025-08-20T02:35:47ZengMDPI AGBuildings2075-53092025-06-011513219110.3390/buildings15132191Leveraging Social Media Data to Understand COVID-19 Prevention Measures in Construction: A Machine Learning ApproachEmmanuel B. Boateng0Daniel Oteng1Dan N. O. Bonsu2Vinod Gopaldasani3School of Social Sciences, University of Wollongong, Wollongong, NSW 2500, AustraliaSchool of Architecture and Civil Engineering, The University of Adelaide, Adelaide, SA 5005, AustraliaChemistry and Forensic Sciences, Griffith University, Nathan, QLD 4111, AustraliaSchool of Social Sciences, University of Wollongong, Wollongong, NSW 2500, AustraliaThe COVID-19 pandemic was a particularly challenging time for the construction industry as it experienced significant disruptions to operations, affecting various stakeholders. With various national and international health agencies promoting preventive measures, the construction industry struggled with the implementation of these measures due to the unique nature of the work involved in construction. This study aimed to highlight the ways in which stakeholders in the construction industry interacted and responded to the prescribed preventive measures through social media analysis. Using model-based clustering and structural topic modelling, this study provided insights into the prevalent discussion topics in social media around prevention measures in construction. In addition, sentiment analysis demonstrated interesting polarisation around the topic areas. Four prevalent topics that encapsulated the entirety of the social media data were identified, with two of the topics showing an upward trend, as expected, while the other two topics showed a contrasting downward trend. These findings offer practical value for construction managers and policymakers by revealing the effectiveness of different communication strategies and identifying areas where prevention measures faced resistance or acceptance. The sentiment polarisation patterns (50% positive, 40% negative) provide actionable insights for developing more targeted engagement approaches, while the topic evolution trends inform the timing and focus of safety communications. Construction organisations can leverage these insights to improve workplace safety protocols and enhance stakeholder buy-in for future health initiatives. This study lays the foundation for future studies to investigate the connections between the prevalent prevention and the interrelated dynamics within the conversation regarding COVID-19 prevention strategies in the construction sector.https://www.mdpi.com/2075-5309/15/13/2191coronavirussocial mediaconstruction industrymachine learningtopic modelling |
| spellingShingle | Emmanuel B. Boateng Daniel Oteng Dan N. O. Bonsu Vinod Gopaldasani Leveraging Social Media Data to Understand COVID-19 Prevention Measures in Construction: A Machine Learning Approach Buildings coronavirus social media construction industry machine learning topic modelling |
| title | Leveraging Social Media Data to Understand COVID-19 Prevention Measures in Construction: A Machine Learning Approach |
| title_full | Leveraging Social Media Data to Understand COVID-19 Prevention Measures in Construction: A Machine Learning Approach |
| title_fullStr | Leveraging Social Media Data to Understand COVID-19 Prevention Measures in Construction: A Machine Learning Approach |
| title_full_unstemmed | Leveraging Social Media Data to Understand COVID-19 Prevention Measures in Construction: A Machine Learning Approach |
| title_short | Leveraging Social Media Data to Understand COVID-19 Prevention Measures in Construction: A Machine Learning Approach |
| title_sort | leveraging social media data to understand covid 19 prevention measures in construction a machine learning approach |
| topic | coronavirus social media construction industry machine learning topic modelling |
| url | https://www.mdpi.com/2075-5309/15/13/2191 |
| work_keys_str_mv | AT emmanuelbboateng leveragingsocialmediadatatounderstandcovid19preventionmeasuresinconstructionamachinelearningapproach AT danieloteng leveragingsocialmediadatatounderstandcovid19preventionmeasuresinconstructionamachinelearningapproach AT dannobonsu leveragingsocialmediadatatounderstandcovid19preventionmeasuresinconstructionamachinelearningapproach AT vinodgopaldasani leveragingsocialmediadatatounderstandcovid19preventionmeasuresinconstructionamachinelearningapproach |