October meetup: A Shiny app for automatic coding of short texts (shinyReCoR)

The Automatic Text Response Coder (ReCo; https://www.reco.science/) is a tool for automatic coding of short texts; for example, responses administered in tests that are scored as either correct or incorrect. The method is based on word embeddings / vector space models (such as Latent Semantic Analysis) and classification, thus borrowing concepts from natural language processing and machine learning, and implemented as a Shiny app (shinyReCoR).

Fabian Zehner, post-doctoral researcher, the author of ReCo and a meetup member from the very beginning, will introduce the method and the Shiny app. There will be plenty of examples for us to explore and play. Please note: ❗We highly recommend installing the app on your device prior to the meetup to make most of it.❗ Below is a description on how to do this.

As always, we welcome useRs at all levels of expertise interested to learn about text processing, Shiny, and R in general.

=== SETUP ===
Video: https://www.reco.science/training-resources-setup-0-2-0/

• Have R installed (at least version 4.0.0)
• RStudio or a modern browser do for working with the package.
• Windows users will need to have RTools installed: https://cran.r-project.org/bin/windows/Rtools/
• download and install shinyReCoR
  ** by typing source("https://www.reco.science/latest") into your R console (RStudio or R will do; if problems occur in RStudio, try R).
  ** Alternatively, you can download the package here and install it manually: https://www.reco.science/download/0.2.0-horizon/shinyReCoR_0.2.0.tar.gz (note that you then need to install dependencies manually too)
  ** Please note that the installation (with the package's many package dependencies) can take a while, depending on your R package library.

=== TEST, START, DEMO DATA SETS ===

• After the setup reports on installation success, you can start the app by calling shinyReCoR::recoApp() in the R console (either RStudio or R)
• a window should appear that welcomes you to the app

Further information (also on troubleshooting) at https://www.reco.science and https://www.reco.science/download-overview

Please feel free to reach out to us: info@reco.science | fabian.zehner@dipf.de