Rating urban transport services quality using a sentiment analysis approach

Abstract

Urban transport systems are recognized as one of the most relevant pillars in the operation of a city, as well as in the quality of service they provide to its inhabitants and visitors. Today, transport network management and control systems play a key role in ensuring their operation, and planning new services in accordance with some specific requirements. Some of these services provide means to users for expressing their opinions about the services they used. When properly treated and analyzed, the opinions reveal very relevant information, both positive and negative, about the operation of public transport companies, which can be used to identify risk situations, malfunctions, or emerging threats, among other things. From the identification of these situations, transport companies can improve their image, attenuating or eliminating such critical and undesirable situations and, consequently, increasing the quality of the service they use and their earnings. In this paper, we present and discuss a system for analyzing complaints about services provided by a public transport company, in order to identify sentiments expressed in complaint texts and establish a complaint-index reflecting the negativity level of the complaint.