[TMP-075] Matching the Right People! Creating a functional demonstrator for online matching of people and expertise for innovation

Contact person: Sebastian Feger (sebastian.feger@um.ifi.lmu.de)

Internal Partners:

1. LMU, Sebastian Feger, sebastian.feger@um.ifi.lmu.de
2. GE, Jan Alpmann, alpmann@german-entrepreneurship.de

External Partners:

1. European Laboratory for Learning and Intelligent Systems (ELLIS)
2. ETH Zurich AI center

Connecting the right people with the right expertise is essential for innovation, but finding suitable collaborators can be challenging. In this micro project, we explored a novel approach that leverages publicly available performance data (e.g., Github profiles) to automatically match people and expertise for innovation. Building on such data, we developed a recommender system that suggests potential collaborators based on their demonstrated expertise.

The project builds on the previous two micro projects, which identified a set of functional questions to match people for innovation and implemented a prototype for creating user profiles based on those questions. In this third micro project, we developed a demonstrator that creates recommendations for collaboration based on the automated collection and analysis of GitHub data. In contrast to the previous micro projects, the approach focuses on demonstrated expertise, rather than relying on self-reported knowledge, which can be inaccurate or incomplete.

Results Summary

To develop the demonstrator, we used the GitHub API to retrieve information about users’ activities and built a recommender system that suggests potential collaborators based on their GitHub statistics. We also created a user interface for founders to find potential co-founders based on the expertise shown. This system offers a reliable and objective way of matching people with the right expertise, which can be especially valuable in high-tech environments where expertise is critical for success.

We evaluated the effectiveness and usability of the developed recommender system and user interface through a series of interviews and online studies. First, we interviewed domain experts to establish the requirements and important features of the model. Second, we interviewed potential founders to evaluate the general approach and gather feedback on the recommender system. Finally, we conducted an online study to evaluate the usability of the user interface. The study involved a group of participants using the interface to find potential collaborators and provide feedback on their experience. Overall, the evaluations provided valuable insights and feedback for improving the system’s effectiveness and usability. We plan to incorporate the feedback from the evaluations to further improve the system and enhance its ability to match people and expertise for innovation in the future. The user-centered design process benefited from close collaboration and interaction with experts working at external partners ELLIS and the ETH Zurich AI Center. The long-term goal is to help connect experts and entrepreneurs across physical boundaries, creating a vibrant and agile high-tech environment on a European scale.

Tangible Outcomes

1. Functional prototype with real-world data from github: http://143.42.16.26:3000/