Understanding User Experience and Satisfaction with Urban Infrastructure through Text Mining of Civil Complaint Data

Abstract

With the increase in public concern about the aging of urban infrastructure and the associated risk of safety accidents, it is important to maintain the safety and serviceability of urban infrastructure in accordance with user satisfaction. Although many studies have attempted to consider user experience and satisfaction based on user surveys and civil complaint data analysis, they have had difficulty in identifying user dissatisfaction factors where users feel unsafe or uncomfortable while using the infrastructure. The main purpose of the research presented here is to understand user experience and satisfaction with urban infrastructure by text mining self-written civil complaint data. To achieve this objective, the researchers adopted the following procedures: (1) development of a civil complaint thesaurus for the text mining of civil complaint data; (2) text preprocessing of civil complaint data by using the thesaurus; and (3) keyword extraction and recognition of the relationships between the keywords to explore user-experience factors related to urban infrastructure. The research team used 2,945 bridge complaint data records and 404 tunnel complaint data records in text format from the Korean Safety e-Report database. From the collected data, the researchers developed a civil complaint thesaurus with 47 semantic relationships between words, such as Korean compound words, synonyms, and hypernym– hyponyms. As a result of keyword extraction, “breakage,” “accident,” and “road” for bridge complaints, and “entrance,” “accident,” and “breakage” for tunnel complaints were the selected words representing user experiences, and were visualized in a tag cloud. Also, critical user-experience factors such as unsafe or uncomfortable situations on bridge roads (e.g., “breakage,” “construction,” and “pothole”), and dissatisfaction factors at tunnel entrances (e.g., “streetlight,” “view,” and “sign”) were explored using semantic network analysis. The outcome of this research will contribute to identifying user-experience factors from civil complaint data and improving the safety and serviceability of urban infrastructure by considering user experience and satisfaction in infrastructure maintenance practices.