A Case Study Documenting the Development of Job Assist: A Speech Based Job Search System for Individuals with Visual Impairments

Abstract

Web accessibility is an improving issue but still an issue nonetheless. The Job Assist system documented within this report addresses the accessibility issue by enabling users with visual impairments to access job and salary information using spoken dialogue interaction through a standalone desktop application. This case study provides information on our practical experience developing the system as well as our experience with Microsoft’s Windows Speech API, Bing Speech API and Language Understanding Intelligent Service (LUIS). In our project we were most severely challenged in our attempt to make our system usable by two groups of users, native English speakers and non-native English speakers of Indian origin. We argue that the solutions that we employed to address these challenges may be of value to other practitioners in similar contexts.

The accessibility of administrative processes

“…There is notable potential for interfaces other than forms to be designed to create more accessible processes and reduce the burdens stemming from limited channels for communication and input, or possibilities for errors or misinterpretations, associated with these interactions. For example, Brinckley et al describe a voice-based interface for job searching, created as an alternative for those with visual impairments [5].…”