Good news everyone! The Monassis practice service is now at version 1.0 and running as part of
Siyavula's
Everything Maths and
Everything Science textbook sites. This actually happened 3 days ago, but I was too exhausted to even think about writing a post.
If you'd like to try out the software, you can register as a user on
everythingmaths.co.za or
everythingscience.co.za, which will get you a free 30-day trial subscription to the service. Note that this is currently a paid-for service that supplements Siyavula's free,
CC BY textbooks, so if you want to use it for more than 30 days, you will have to sign up for a paid account.
What's under the hood?
Templated questions: Each question that you see is actually generated from a Python and XML-based template in a
MVC framework. Numbers, names, variables and a little bit of problem structure are generated randomly each time that you reload a template. We currently have about 350 templates, which might not sound like a lot but each templates equates to a bunch of individual exercises.
Cool maths parser: Maths problems requiring symbolic responses -- like factorised or expanded polynomials, inequalities, intervals on the real line, proper fractions, sequences and sets -- are all parsed from a simple calculator-like notation; and compared to a model solution.
Instant feedback: This is a practice service, which means that you get a detailed solution to each exercise once you've submitted your response.
Dashboard: Simple feedback on your progress with different types of exercises. Exercises are organised by chapter in the Everything Maths and Science textbooks and users see a breakdown of their progress per chapter.
What's next?
Gathering data: You can't make an omelette without eggs and you can't do inference without data. Now that the service is live, we'll be encouraging people to sign up with the aim of getting as broad coverage as possible of school-age students in the country.
Diagnostics: We're already gathering usage data from the textbook websites, giving insight into how our books get used around the country. This will be extended to mining user data from the practice service for a couple of purposes:
- helping individual users identify weaknesses and helping them find educational resources to remedy those weaknesses;
- discovering correlations between exercises across grades and subjects, to better understand how students can learn difficult concepts;
- reporting on the state of education in South Africa, to inform decision about educational programmes or interventions.
Scalability: We're aiming to get a lot of people to sign up and we'll need a bit more work on the hardware and software to handle the load.
More templates: Work continues to get good coverage of all the chapters in all our current and future books. Making templates takes time, effort, technical skill and good exercise design. Our crack team of technical content editors is on the job.
Mobile interface: Most of our target audience use mobile phones to access online content and the service will soon be re-styled to accommodate them.