Overview and Problem

HM Hochschule München University of Applied Sciences lacks a central point of contact that can provide information on all areas of study. Students at the university currently have to search through various websites and platforms for information. There is no central place that can provide information 24/7 and that can quickly provide answers to questions regarding studies. A chatbot could be the solution to this problem.

Procedure

The challenge ran under the main title "Chatbot". A total of four student teams were commissioned to create chatbots, each for a topic related to the university. The results of a survey and the "empathy map" created by the students made it clear that information about deadlines (e.g. the period for registering for exams, the duration of the semester or the re-registration period), but also the search for sample exams, are particularly relevant for the students. For this reason, the second team was given the task of designing a chatbot to help with questions on these topics.

In the further course of the challenge, several use cases were identified and described by means of user stories. In three sprints, the students also specified so-called "conversation flows" and developed the prototype of the chatbot "Cubie" based on AWS Lex. At the end of the challenge, the prototype was tested and integrated into the university's learning platform (Moodle). The student teams were supported by Amazon Web Servics (AWS).

Innovation in action

The chatbot Cubie works as follows: First, it informs the user about the possible topics for which information can be provided. The user then says, for example, that he or she is looking for old exams. Cubie then asks for which course the sample exam is wanted, saves the answer and displays existing sample exams. These appear as a clickable link to the file in the chat window in Moodle. In addition, an icon indicates that a corresponding file has been found.

Next steps

The prototype needs to be technically refined in the future.