| Science and Research

Chatbot colleagues: Better information for students

Chatbot colleagues: Better information for students

Chatbots can do more than relieve customer support representatives in the service area. For example, they can be used to optimize ranges of information. A project by the European Research Center of Information Systems (ERCIS) at the University of Münster shows what this could look like.

The University of Münster (Westfälische Wilhelms-Universität (WWU) Münster) is one of the largest universities in the state, with 15 faculties and around 45,000 students. The institution, which is rich in tradition and whose history dates back to the 16th century, has been at the scientific pulse of the times for more than 400 years. A chatbot project is currently taking place there; the project was initiated by the School of Business and Economics and is supported by Arvato CRM Solutions. A project team, consisting of five members from the Business Information Technology department, is developing a chatbot for the examination office.

The project started in April, in the summer term of 2018, and will run for five months. The bot, named PAMBOT, uses the natural speech comprehension of IBM’s Watson and thus offers the 6,000 business and economics students in Münster an easy, convenient way to access information about the department’s 376 exams and seminars. Among other things, the chatbot can answer questions like: Where and when do my exams take place? Which exams have to be taken in the sixth semester of business studies? Which exams and seminars belong to the “Corporate Accounting II” course? Until now, students had to collect this information in a rather complicated way from the department’s homepage.

Already in the first week of the beta test phase, more than 34 percent of visitors to the homepage accessed the chatbot. And the users didn’t stop at asking the bot just one question. On average, they asked three questions per interaction and so, small dialogs evolved in the early stages of development. The solution is now to be further expanded until it can finally replace the current range of information completely.

Close cooperation between the university and the company

The chatbot is based upon the “Arvato Conversational AI” platform by Arvato CRM Solutions. The students are working closely with teams from Arvato during the implementation. For this, they are relying on an agile approach with the scrum methodology. The chatbot’s prototype was successfully completed at the end of August. PAMBOT is now permanently available to students, and its use is being further evaluated. But in the meantime, important insights have already been found:

  • There are not only interesting cases of applications for chatbots in the service areas. Fundamentally, they can be deployed anywhere, which has to do with the simplification or optimization of communications in the broadest sense.
  • Performant technology is a basic requirement for a chatbot. It is crucial for its success, but also for user acceptance, and this comes much more from the structure of the dialogs, the way the bot replies.
  • Less is often more. The School of Business and Economics’ chatbot offers no technical features or gadgets. Instead, it is accurately geared towards the needs of the target group and is thus all the more effective.

“The project is another success drawn from the close cooperation between the University of Münster and Arvato CRM Solutions and it shows the clear added value for all those involved,” says Prof. Jörg Becker from the Department of Information Systems at the University of Münster. Becker continues: “The University of Münster has a long tradition of working with partners from the business world in project seminars. As a result, students can closely combine theory with practice – the ideal entry into a successful professional career. The PAMBOT project shows how the life of students can be quickly changed by innovations, such as chatbots.”

Author: Editorial team Future. Customer.
Image: © weedezign – AdobeStock

Tags for this article Artificial Intelligence (85) Chatbots (37) Digitization (167)


team members






clients worldwide

We're global and local. So you can be close

We're global and local.

We speak your language. And we’d love to hear from you