COVID-19 impact on end-user's maintenance requests. A text mining approach

COVID-19 pandemic changed our way of working, limiting the usual physical attendance of working spaces. Despite the drastic reduction in the number of daily users due to the pandemic restrictions, working buildings were often kept open to provide services to internal and external users. Pandemic obliged to change operation and maintenance (O&M) plans, due to the increase of ventilation requirements and the reduction of other types of services, with a strong impact on cost and management. Now the pandemic is reducing its effects and is time to question the future asset of buildings’ O&M plans, based on the pandemic lesson. Data collected by Computerized Maintenance Management Systems (CMMS) during COVID-19 then become an important source of understanding the future management of working places. End-users’ maintenance requests are usually expressed by natural language, then a text mining approach can be a useful tool to discover hidden knowledge from unstructured data stored in CMMS. This study applies text mining methods, including sentiment analysis, to the field of building maintenance, with the scope to evaluate how COVID-19 changed some aspects of the facility management process, including users’ perception.