talkr offers a set of convenience functions for quality control, visualisation and analysis of conversational data.
It provides support for at least two common needs of people working with conversational corpora:
(1) quality control by offering rapid insights into the nature, timing, and quality of time-aligned annotations in a conversational corpus; and
(2) compelling visualisations by offering a range of plotting functions that play well with ggplot and the tidyverse.
Note: talkr plays well with its companion python package scikit-talk, which is focused more on processing conversational data and readying it for downstream analysis.
You can install the development version of talkr from GitHub with:
# install.packages("devtools")
devtools::install_github("elpaco-escience/talkr", build_vignettes = TRUE)talkr has been submitted to CRAN, and will soon be installable with:
install.packages("talkr")Two workflows for talkr are described in the workflows vignette.
You can also access the vignette from RStudio, by running:
vignette("workflows", package = "talkr")Questions? Comments? They are more than welcome! This package is under active development and still has some rough edges. Interact with us in the issues if you have any questions or suggestions.