(Back to publications)

AI resources

I am interested in finding ways in which Articial Intelligence and Machine Learning can support the practice of mathematics education research and influence the development of theories. On this page, I share my research findings and the resources used for producing them.

Automated analysis of mathematical content in teacher talk

Note: Data was translated automatically through Google Translate and will likely not produce good results when used for training a model, as automated translation likely mistranslates technical mathematical terms and natural language with mathematical connotations.

  • Source code of the model training (last updated 06.06.2019) (Download source code as jupyter Notebook(recommended) or as text file )
  • The trained model as a .h5 file. It can be loaded through the keras load_model() function (last updated 06.06.2019) Download model
  • Teacher utterances labelled for mathematical content - original German (last updated 06.06.2019) (Download original German or translated to English)
  • The predictions of the trained model for the test set (last updated 06.06.2019) (Download original German and translated)