In order to identify trends and changes in the field of HCI, we used the full-texts of the papers of the HCII conferences from 2007 to 2017 in a text-mining approach. From a set of approx. 7500 documents we looked at word frequencies and topic modelling using latent dirichlet allocation (LDA) in order to detect changes and trends. We identified 50 topics using the LDA model. We found that the topics around social aspects, gamification and datafication play an increasing role. We find evidence for this in both LDA and word frequencies. We qualitatively asses the topic models using our own publications and find a high match of detected topics and our ground truth.