Learning building occupants’ indoor environmental quality complaints and dissatisfaction from text-mining Booking.com reviews in the United States
Nan Ma, Qi Zhang, Fabrício Murai, William W. Braham, Holly Samuelson
Abstract
Traditional post occupancy evaluation (POE) poses a challenge to the comprehensive knowledge of occupant dissatisfaction with indoor environmental quality (IEQ) due to the nature of the pre-defined questionnaire structure and time-consuming data collection procedures. At the same time, the fact that “humans are effective sensors” and are inclined to report their impressions can be leveraged to collect data for a substantial number of buildings. This paper presents a text-mining approach to reveal the relationships between environmental conditions and occupant dissatisfaction in large populations. With a set of 1.2 million Booking.com open-ended reviews (i.e., 71,665 IEQ reviews) of temporary residences (e.g., hotels), we identified the prevalence of IEQ complaints, investigated the climate distributions and seasonal trends of IEQ complaints, and quantified the social benefits of IEQ in economic values. Indoor Air Quality (IAQ) complaints make up the majority of dissatisfaction ratings in all climate zones and IAQ-related reviews also have the worst sentiment scores, followed by visual, acoustic, and thermal discomfort. Windows are perceived as the common source that matters for all four categories of IEQ dissatisfaction. Our analysis also found that the IEQ value is correlated with increased rental costs when the temporary residences have satisfactory IEQ conditions. Our study demonstrates that IEQ satisfaction can be a key design driver and therefore, should be considered in building design, management, and real estate finance.