Towards Occupant-Centric Facility Maintenance Management: Automated Classification of Occupant Feedback Using NLP

Location-specific occupant feedback is crucial information for facility management since it reflects occupant satisfaction on the facility and informs corrective maintenance. However, the current facility maintenance management (FMM) system cannot effectively collect location-specific occupant feedback, resulting in the inefficiency and ineffectiveness of FMM. This research explores a mobile social network approach for location-specific feedback collection in FMM. Specifically, this research develops a natural language processing (NLP)-based machine learning algorithm to classify the occupant feedback collected from the social network application automatically. In turn, the proposed approach could facilitate occupants’ participation and occupant feedback solicitation and assist facility management (FM) personnel in locating and navigating the reporting issues. The Waikato Environment for Knowledge Analysis (Weka) is utilized for training and testing machine learning algorithms based on historical records. The trained model can retrieve the essential information to facilitate FM decision-making based on occupants’ input data (e.g., location and textual information). The technical feasibility of mobile social network applications to report FMM feedback and concerns was demonstrated using a case study. This research contributes to the body of knowledge by an NLP-based model that automatically processes the location-specific occupant feedback for FMM. Future work will focus on developing a prototype application for FMM based on our findings.