Teaching Tip: An On Demand Text-to-Speech Agent
George H. Bodnar · The Journal of Information and Systems in Education · 2008
This Teaching Tip provides a Microsoft .NET 2.0 framework application that uses Microsoft Agent[R] for on demand text- to speech (TTS) capability. One opens the program, then pastes selected text into a textbox to have Microsoft Agent[R] Merlin read, i.e., speak the text. One may also open text files directly into the application. One can use the application in a multimedia classroom environment to hold and enhance student interest with a variety of voices and also to provide some technological assistance when it is desired to have text read-aloud. Keywords: Multimedia Classroom, Microsoft Agent[R], .NET Framework, Text-to Speech Output 1. INTRODUCTION Contemporary multimedia classrooms have sound capability. This permits the use of text-to speech (TTS) output technologies. A primary objective is to hold and enhance student interest with a variety of voices--the human lecturer/narrator and one or more TTS voices. This paper provides a Microsoft .NET framework 2.0 application that uses Microsoft Agent[R] for on demand TTS capability. 1.1 Microsoft Agent[R] Microsoft Agent[R] is a set of programmable software services that support the presentation of interactive animated characters within the Microsoft Windows[R] interface. Microsoft Agent[R] enables software developers to incorporate a new form of user interaction, known as conversational interfaces, that leverages natural aspects of human social communication. Microsoft provides speech output (text-to-speech or TTS) engines that one can use with Microsoft Agent[R]. 1.2 Combining Human Speech and TTS A number of applications and studies have addressed this topic. Ng S. T. Chong, Panrit Tosukhowong, and Masao Sakauchi discuss how the use of TTS (text-to-speech) synthesis is becoming increasingly attractive as a supplement or even as an alternative to human narration, and they present the results of a study that examines the combined use of TTS and human narration on Web lecture presentations (Chong, et al., 2002). They conclude that their study validates their intuitions about the benefits of TTS technology in Web lectures. TTS demonstrates merits where the entire lecture is narrated using TTS or where the lecture contains both speech synthesis and human voice. While their project concerns Web based presentations, both it and the enclosed application possess the basic feature of augmenting human narration with TTS to capture audience attention and to hold their interest. In a paper titled Use of Interactive Multimedia to improve the lecturing experience, Clive Chandler presents a review of the large body of work that addresses the topic of using multimedia in the field of education (Chandler, 2003). The traditional lecture is a basic tool in transferring information. The paper discusses the development of an agent assisted lecture process and an analysis of its effects. Chandler states On the whole there would appear to be much mileage in the use of these characters in an interactive way to enhance the lecturing process. A potential drawback is that the lecture may become too much of a performance. 1.3 The TTS Application AgentTalk The application, titled AgentTalk, activates Merlin, a Microsoft Agent[R], and provides basic TTS capability by having Merlin read selected text. The operation of the application is discussed in a subsequent section. The design goal of the application is to provide basic TTS capability in direct fashion. The .NET Framework makes it possible to run an application such as AgentTalk directly from a personal storage device on demand. There is no need to have the program previously authorized or installed by an administrative function on the target computer. The .NET Framework handles the details of locating and loading the components an application needs. This is known as xcopy deployment. …