Speech Recognition Technology in K–12 STEM-Driven Computer Science Education

This chapter presents a methodology for introducing speech recognition (SR) technology in K–12 STEM-driven CS education. Three tasks (Task 1, Task 2, and Task 3) are under investigation, though to a different extent. Task 1 concerns the basic concepts of SR with as many details as possible for learning. Task 2 is about the voice spectral analysis. We outline it without implementation, i.e., as future work. Both are independent of using robotics in education. Task 3 is robotics-dependent. It concerns the processes of controlling the robot’s actions by voice commands. In general, this task is about human–machine interaction. From the STEM context, it is a significant real-world task. In the past, there have been intensive efforts to improve this interaction. The human–machine communication by voice is the old dream of computer scientists, which is becoming a reality now. Therefore, we present Task 3 with as many details as possible. The introduced methodology includes the following topics: (i) description of the basic idea with the motivating scenarios for solving Task 3; (ii) background, i.e., definitions of SR basic terms, their properties, and relationships; (iii) research methodology presented through the generic process model; (iv) extended Smart Learning Environment (SLE) for SR tasks. The experimental part includes two case studies with results. Case Study 1 presents the results of solving Task 1 with details in Appendix 1. Case Study 2 presents the results of solving Task 3 with a detailed description in Appendix 3. Appendix 2 gives a template for solving Task 2.