Significant recognition has been bestowed upon the Master in Entrepreneurship students Lukas Schürmann and Raphael Derstappen of WHU – Otto Beisheim School of Management: Both have been invited to present the research results of their respective master’s theses at internationally renowned conferences. In order to do so, they had to go through a competitive selection process in which they competed with research projects developed by doctoral students and professors. Lukas Schürmann will present the findings of his thesis "Using text-based machine learning algorithms to analyze entrepreneurial ecosystems" at the International Product Development Conference. Meanwhile, Raphael Derstappen will present his findings at the Strategic Management Conference.
Better alignment of research and start-ups in artificial intelligence
For their research, both students used advanced natural language processing methods to investigate specific ecosystems for businesses. Lukas Schürmann evaluated data from 21,000 international start-ups for his thesis. From their activities, he was able to identify a total of 50 different application areas for artificial intelligence (AI). Based on these findings, he analyzed data from more than 2,000 public AI research projects. In comparing them, he found that the uses for AI on which German institutes primarily conduct research and for which they receive funding from the federal government correspond only to a small extent with the AI topics in which the German start-up scene is interested. This is a remarkable result, given that the German government is spending billions on funding its AI strategy until 2025.
A new algorithm through testing in the cybersecurity ecosystem
Like Lukas Schürmann, Raphael Derstappen used advanced machine learning technologies. He conducted a systematic comparison of different algorithms of topic models and illustrated to what extent they are suitable to perform management research. In doing so, he tested the models on enterprise data from start-ups in the cybersecurity ecosystem. From the comparative data and its accuracy, Derstappen was able to create a new algorithm for topic models which combines the advantages of all previous models and minimizes their weaknesses. Both WHU students were invited to the respective conferences for reasons extending far beyond their knowledge of advanced machine learning technologies.