Abstract
Aims: This article reports the development and the experimental results of the acceptability testing and comparison of speech and dual tone multi frequency (DTMF) for interacting with a speech application for carrying out banking transactions. Study Design: A within-subjects experiment where each of the participants tested both the DTMF only and speech only modalities was carried out. Each one filled a questionnaire after testing each modality Place and Duration of Study: Department of Computer and Information Sciences, Covenant University, Ota, Nigeria, between May 2013 and June 2013. Methodology: Voice Objects 11 was used as the development platform. Voxeo Prophecy 9 was used as the implementation platform while MySQL 5.5 was used as the database management system for storing each customer's account details. X-Lite soft phone 3.3 was used for testing the system. 50 undergraduates of Covenant University, Ota, Nigeria participated in the evaluation. Each one tested the system using DTMF and speech separately. After each round of the tests, each one filled a questionnaire for the modality tested in a bid to measure the acceptability and user satisfaction with the modality. Results: The entire system’s satisfaction ratings for DTMF were significantly higher (M=36.18) than for speech (M=34.76), t(49)= -1.46, tcrit =2.0, P=.05. For modality evaluation, speech was more satisfying (M=12.22) than DTMF (M=12.18), t(49)= 0.14, tcrit =2.0, P=.05. For modality entertainment, both speech and DTMF were rated equal (M=13.94), t(49)=0, tcrit =2.0, P=.05. For modality naturalness, speech was more natural (M=10.2) than DTMF (M=9.44), t(49)=- 2.41, tcrit =2.0, P=.05. 52% of the subjects chose DTMF modality, whereas 48% chose speech (p = .05). Conclusion: Dialogue systems are a widely acceptable technology for carrying out banking transactions in Nigeria. It will provide a cost-effective and easily accessible means of carrying out banking transactions
Highlights
Speech technology incorporating Automated Speech Recognition (ASR) and Text-to-Speech (TTS) enables humans to interact with electronic devices through human language
The null hypothesis is that the mean difference between dual tone multi frequency (DTMF) and speech user satisfaction is zero
The null hypothesis is rejected, and the finding indicates that DTMF was more satisfying than speech
Summary
Speech technology incorporating Automated Speech Recognition (ASR) and Text-to-Speech (TTS) enables humans to interact with electronic devices through human language. Speech technology can be used to bridge the digital divide [1] and is capable of providing widespread access to services by the people, by exploiting the ubiquitous and widespread availability of mobile phones It has the advantage of being usable by the non-literates in Africa and suitable for the visually impaired. The dual tone multi frequency (DTMF) part of the technology is useful for people with speech disorder since it gives them the option of responding to speech applications by pressing the keys on the telephone keypad instead of a spoken response. It can be implemented in the various indigenous languages of Africa and can provide eyes-free interaction. It can serve as a great aid for the physically challenged [2]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
