Abstract:

Universities can as little ignore the ever-faster pace of change in their environment as companies can. To anticipate the increase in the quantity of knowledge and the increasing competition in the higher education sector, universities can access their own knowledge base. This paper aims to show that universities can use machine learning to uncover potentials in their own knowledge base and use them for curriculum development. The final theses written at a German university are examined using software based on the Latent Dirichlet Allocation (LDA) topic model. Topic blocks are disclosed that reflect current developments and trends on the market. This paper describes the process and essential features in the preparation of the corpus and the derivation of the thematic areas. The added value of this method is pointed out and subsequently discussed using examples. This results in an innovative approach to future-oriented curriculum development. The study presented here gives only a small insight into the fundamental changes which artificial intelligence will imply for all of us in the future.