scikit-talk user documentation

Scikit-talk will be an open-source toolkit for processing collections of real-world conversational speech in Python. The toolkit aims to facilitate the exploration of large collections of transcriptions and annotations of conversational interaction, specifically tailored to applications in speech processing and Conversational AI.

Scikit-talk is under active development, and currently includes a preprocessor module that allows the import of several common transcription formats.

You can read the full paper here: Liesenfeld et al. (2021)

Contents:

• Installation
• Getting started with scikit-talk
  • Parsing a .cha file
  • Parsing an .eaf file
• Analyzing turn-taking dynamics
• The Corpus object
• Storing and retrieving Conversation and Corpus objects
  • json
  • csv
• API Reference
  • sktalk