3. september 2010

Romanian students from DIKU develop an algorithm library

Paula Petcu and Radu Drăgușin, two Romanian 2nd-year graduate students who study at DIKU, got their paper "PyAlg: An Algorithm Learning Platform" accepted to the 5th International Conference on Virtual Learning.

Paula Petcu tells about the project:

"This paper was produced in connection with our work on the two-block project-course 'Development Studio'. The course responsible, Jyrki Katajainen, presented to us the difficulties of the first year students with learning algorithms and proposed an algorithm library to improve their learning experience. We saw this as a great opportunity to put into practice some ideas we had since we first started learning algorithms, years ago".

"In Romania, as well as here in Denmark, many first-year students have a bad experience with the algorithms course. Some find it difficult, others find it boring, but nevertheless this is one of the fundamental courses in our field of study".

"Jyrki was open to suggestions and encouraged us to put our ideas into practice. We used our own experience with learning algorithms in order to bounce ideas with him".

"In the end, we created a tool we would have wished was available when we were learning algorithms. We hope that our work will improve the experience of students learning and teachers teaching algorithms".

"After the course completion, we decided to disseminate our work in a conference on e-learning. Our paper is now accepted for the 5th Conference on Virtual Learning that will take place at the end of October this year". 

About the paper: 

PyAlg: An Algorithm Learning Platform 

By Radu Drăgușin and Paula Petcu


The course on algorithms and data structures is a fundamental course for those studying computer science. Thus, a huge amount of teaching material covering this topic has been developed over the years: from books on algorithms and data structures and articles on algorithm learning, to algorithm libraries and algorithm-visualization tools. However, our study of previous work shows the lack of a unitary solution that covers essential aspects of algorithm learning: algorithm libraries, visualization, and benchmarking. The aim of this project was to develop a unified interface covering those aspects. The developed tool, named PyAlg, is a learning platform that can be used in studying, teaching, and analysing algorithms. It was originally designed for the students following the undergraduate course on algorithms and data structures in the Department of Computer Science at the University of Copenhagen, where it was used in one of the programming assignments. The target audience for the latest version of the developed product is wider, though the objective remains the same: helping students in the process of understanding, learning, and analysing algorithms.