AI-supported Natural Language Processing in project management – capabilities and research agenda

Helge Nuhn, Alfred Oswald, Agnetha Flore, Lang Rüdiger
AI-based natural language processing (NLP) models show remarkable performance in tasks like question answering or text generation in general. We argue that recent NLP-AI models will play a major role in the transformation of our societies, an endeavor mainly shaped by projects, project managers and project teams. We present the results of an experiment, in which we fed state-of-the-art NLP models like GPT-3 project management-related questions and had an expert team rate the maturity of the answers. Results indicate that the size of the model and the text corpora seem to make substantial difference to the performance. The best model seems to be able to answer most project management-related questions convincingly to the judgement of the expert panel. Attempts to train a model without extensive training data preparation resulted in poorer results. A research agenda is derived and presented.
Dezember / 2022
IPMA Research Conference 2022
Proceedings of the 10th IPMA Research Conference “Value co-creation in the project society”
Zentrum für digitale Innovationen Niedersachsen (sorry - only available in German)