Necessity is the mother of invention, as we know. In the Fall Semester of 2014, Maik Meusel, teaching and research assistant at the Department of Business Administration at UZH, faced a Herculean task: The case studies of 170 students from a statistics course needed to be marked. In order to be more efficient, he started to program automated answers for specific questions.
What began with a few prototypes in 2014 grew into a sophisticated learning management system over the years. Today, SYLVA – Latin for forest – is a web platform on which instructors can create courses and set exams, as well as assess and administrate them through automated processes. Around 3,000 students have put SYLVA to the test so far, many of them from UZH. Teachers using SYLVA include staff at UZH’s Faculty of Business, Economics and Informatics. The learning platform is also used in the US, the UK, Greece and Germany – mainly at business schools.
To establish SYLVA on the market, Maik Meusel founded a start-up with Karl Schmedders, former UZH professor of quantitative business administration, and Jessica Sudo, a business administration researcher at UZH.
This was a courageous step, as online learning and assessment platforms are in abundance. But most of them don’t offer what SYLVA provides. Many platforms, for example, only permit simple question formats such as multiple choice. Furthermore, these learning management systems often feature very limited marking options and operate with predefined model solutions. “SYLVA, in contrast, encompasses many different question types, including open questions and interactive exercises,” explains Meusel. This is particularly important for academic subjects such as statistics, mathematics and economics where the use of formulas and models is essential.
Software from Siri
A core element of SYLVA is the automated function for individualizing and marking exams (see box). The smart technology behind it is based on the expertise of a big player in the cloud software sector: Wolfram Research. The American firm provided the algorithms for the voice-assisted systems of Siri and Alexa, among others. Maik Meusel met Stephen Wolfram, the founder and CEO of Wolfram Research, at a conference. “I suggested his software could also be used for automating exams,” says Meusel. His proposal was well received and Meusel has collaborated with Wolfram Research since he started developing SYLVA.
Clear learning outcomes
The advantages of SYLVA are apparent: The simplified marking and evaluation process saves instructors several hundred hours a year – time which can be used for advising and supporting students. The benefit to learners is two-fold. The instructor has more time for them, while they also profit from SYLVA’s automated evaluation. This enables the students to access numerous self-tests and receive immediate feedback concerning their level of knowledge. Thanks to the automated parameterization (see box), teaching staff can even create different versions of the same question, allowing students to complete ever-varying exercises during the self-tests. “This significantly accelerates the learning process,” adds Meusel.
SYLVA is predominantly aimed at teachers in the fields of mathematics, computer science, natural sciences, technology and economics. “In the Western world, our market encompasses more than 50 million students,” Meusel confidently declares. Though he does concede that there are many competitors and new learning tools are continually being developed: “There’s currently something of a gold-rush spirit.”
In addition to business schools, the UZH start-up has already gained other interesting clients: Pearson Education, the world’s largest publisher of school and university textbooks, will use SYLVA’s technology for mathematics problems on its own learning platform from this fall. This business concept, selling his own technology to other providers of learning platforms, is something Meusel wants to continue pursuing in the future. The funding program Venture Kick is also convinced of the potential of this young enterprise and has recently decided to back it with 150,000 Swiss francs. “This outside capital helps us to drive forward our digital marketing,” says Meusel. To realize further plans – such as employing permanent staff and renting office space – the start-up is planning a larger financing round towards the end of this year.
How automated parameterization works
“Name a town within 300 kilometers of London that is located outside of Great Britain.” Using an input mask, instructors define the conditions for the answer and specify the relevant number of points. They can also apply various parameters, for example for the distance to London, so that each student receives a different question. Once the student has entered the answer (town), the software calculates whether it fits the defined conditions and automatically issues the evaluation or grade.
The editorial team reserves the right to not publish comments. We will not publish anonymous, defamatory, racist, sexist, otherwise prejudiced, or irrelevant comments. UZH News will also not publish comments with advertising content.