Evaluation of a Natural Speech-Based Application to Increase Transit Utilization

In collaboration with MIT Computer Science & Artificial Intelligence Laboratory or CSAIL, this project will explore the potential of a user friendly, natural speech based information inquiry application as one means of increasing public transit utilization. We suggest that a key challenge to expanding transit ridership is to encourage people who have not used a system to develop familiarity with it. One example is older adults who have lost the ability to use their car or who should reduce their dependence on it. Various factors contribute to the hesitancy of potential new users to try-out public transit options, including uncertainty about routes, schedules, and other basic utilization information. Individuals who are unfamiliar with routine maps and schedules often are unsure of how to begin making sense of these materials. Automated information systems provide one means of accessing such information in an anonymous and potentially non-threatening manner. In particular, the development of next generation natural speech based inquiry systems offer the attraction of obtaining information about the transit system through the familiar conversational method of asking questions. The proposed application could potentially be deployed in various forms (such as phone access, internet or kiosk). The initial phase of the project will focus on gathering strategic information from potential users on implementations they would be most likely to use and the types of information they perceive as being most useful in supporting and/or increasing their use of the transit system. Particular attention will be focused on how demographic factors such as age and gender, as well as technology experience, impact on preferences and likely willingness to consider using various implementations of such systems. The follow-on phases will involve actual development and evaluation of a demonstration application.