Ode to the Automata Tutor

I am slowly emerging from teaching a first year course on topics in Discrete Mathematics to about 250 students at Reykjavik University. (I have one more lecture to deliver after Easter and then the not-so-small matter of over 200 exam papers to grade :-() This is the second discrete mathematics course the students take and it is the second spring semester in a row that I teach it.

As part of the course, I am supposed to cover  the basics of grammars, finite automata and regular expressions. This is a first year course, so I do not cover much of the theory related to these formalisms and their connections. Mostly I expect the students to be able to design grammars, finite automata and regular expressions for some relatively simple languages.

In both editions of my course, which has been followed by about 500 students overall, I have used the AutomataTutor to support my teaching of material related to finite automata and to grade student assignments automatically.

For what it is worth, I strongly encourage my readers to try the tool and to use it in their undergraduate courses. In my experience, the students love to learn DFA and NFA programming using the Automata Tutor and to work on assignments that employ it. The automatic feedback and grading provided by the Automata Tutor are almost magical. (See this paper for a description of how the tool does both.) This is how the construction of finite automata that recognize regular languages should be taught in a modern way! I wish I had similar tools for all the topics I need to cover in that course.

From my perspective (and from that of my TAs), automatic grading is a real bonus. I love to teach, but I really hate to grade a large number of student assignments. Students can be very creative in their solutions and grading them is a very time consuming, haphazard and inconsistent process for any human. The algorithms embodied in the Automata Tutor produce consistent results at the press of a button and the students receive a grade straight away as well as excellent hints on how to improve incorrect solutions.

Thanks to Rajeev Alur, Loris D'Antoni, Sumit Gulwani, Dileep Kini, Mahesh Viswanathan and their co-workers, teaching basic finite-automata theory to hordes of first-year students can now (largely :-)) be done without tears. To boot, the folks at Automata Tutor have always been ready to provide technical help, when that was needed.

I'll keep using the Automata Tutor in my courses and I hope that you will do so too. That is the best way to thank our colleagues for the work they have done and are still doing on that tool.

February 2015 issue of the Bulletin of the EATCS on line

Thanks to the work of Kazuo Iwama, editor in chief of the bulletin, and of his collaborators, the February 2015 issue of the Bulletin of the EATCS is now available on line. You can download the whole issue in PDF from here, if you prefer. As a service to the TCS community, the bulletin continues being open access because of the support of the members of the EATCS, whom I thank wholeheartedly.

Amongst other things, this issue contains contributions by David Eppstein on K-Best Enumeration, Vikraman Arvind on Robust Oracle Machines revisited, Gaudi Taubenfeld on A Closer Look at Concurrent Data Structures and Algorithms (pages 59-82 of the whole issue), Jukka Suomela on Local Coordination and Symmetry Breaking ((pages 83-110 of the whole issue), Andreas Blass on Negative Probability and by Juraj Hromkovic who kicks off the new-look Education Column with a piece entitled Homo Informaticus.

I welcome Stefan Schmid as new editor of the Distributed Computing Column and thank Panagiota Fatourou for her sterling editorial work over many years.

Enjoy it. 

Video made by the female CS student association at Reykjavik University

I learnt a few things about some of the students taking the first-year course I am teaching right now by watching this well-made video.