A project for voice-based metadata creation. Metadata creation is a desirable but often tedious task. As speech recognition has been applied in many industries, we were keen to explore it's potential for map metadata creation. We used the Mozilla's speech recognition API to texualize voice and transforms the text generated to a JSON-LD format. The code on this repostitory contains the prototype and instructions for the user study. We use the Vue framework to realize the application. If you are interested in how we do the voice to text, you can refer to this page. The deployment of the work can be found on heroku.
For more details on this project, please refer to the following publication:
Lai, P.-C. and Degbelo, A. (2021) ‘A comparative study of typing and speech for map metadata creation’, 24th AGILE International Conference on Geographic Information Science (AGILE 2021).
