FAQ (Frequently Asked Questions)

  • Are available the audios?

No. Currently you cannot access audio recordings through the platform. If you need access to any of the audio recordings, contact (adrian.cabedo @ uv.es).

  • How morphological tagging has been done ?

The morphological tagging has been done with Freeling free tagger  (http://nlp.lsi.upc.edu/freeling).

  • Is all corpus correctly tagged?

 75% of the recorded material is correctly tagged. The remaining 25% is currently under review. It should be noted that some of the colloquial Spanish usual phenomena (restarts, hesitations ...) and even possible transcription errors may involve tagging f

  • Is spontaneous all the material collected in the corpus?

90% of the corpus consists of conversations collected from real speech and using secret recording technique. 10% restant belongs to semi-structured conversations where people knew they were being recorded.