The Future of Education
FEHA: An Adaptive Web-Based Front-End Environment to Support Hands-On Training in Parallel Programming
Syunji Yazaki, The University of Electro-Communications (Japan)
Takeshi Kikuchi, Tokyo University of Technology (Japan)
Hideaki Tsuchiya, The University of Electro-Communications (Japan)
Hiroaki Ishihata, Tokyo University of Technology (Japan)
Abstract
In this manuscript, we propose an adaptive web-based programming environment called Front-end Environment for Hands-on Activities (FEHA). Parallel processing technology has become essential in
computer science. Many applications and cloud service infrastructures have been parallelized to achieve better performance and quality of service. However, preparing parallel programming lecture
is a challenge because such programming environment usually requires a computer cluster that include many number of computers, network, software. Some academic organizations have cluster
systems for research projects however utilizing these systems for lectures is difficult because the systems do not have a user-friendly interface for learners. To tackle this problem, we developed FEHA that provides a web-based integrated development environment (Web IDE) for UNIX-based computer systems including computer clusters and
supercomputers. FEHA has some buttons to specify options for parallel computing. FEHA also shows code snippets and hints for programming beginners. FEHA is implemented by using Node.js and MongoDB. FEHA compiles and runs submitted programs from learners on the cluster systems with concise security parameters. FEHA can utilize
almost every UNIX system because FEHA uses only two common UNIX programs: “ssh” and “rsync”. “ssh” is used to invoke any UNIX commands on the cluster. “rsync" is used for incremental file copy between FEHA and the cluster. We used FEHA in two lectures at a university. Students learned fundamental parallel programming methodologies in the lectures. In the first lecture, 108 students submitted 1,658 programs through
FEFA while 116 students submitted 1,648 programs in the second lecture.
We conducted a questionnaire survey to evaluate usability of FEHA by using WUS. WUS is a commonly used usability evaluation scale. WUS evaluates websites on a scale of 1 (poor) to 5 (good).
We also asked additional questions for farther analysis. As a result, students, in average, put 3.3 and 3.2 in the first and second lectures, respectively. We also found that both students that like and
dislike programming gave higher grade to FEHA's characteristics of “reliability” and “easy-to-use”.From the result, we confirmed that FEHA contribute to make the parallel programming lectures more
effectively and easy-to-learn.
