You collected a large number of texts and think it is a good idea to summarize your corpus into several textual sentiment time series, which you ponder could help predicting some variable you are interested in. However, you do not really know how to proceed next… Fortunately, you come across the sentometrics package, which does exactly what you need! Great!

Installation

To install the package from CRAN, simply do:

install.packages("sentometrics")

To install the latest development version of sentometrics (which may contain bugs!), execute:

devtools::install_github("sborms/sentometrics")

Examples

Check out the Examples section. It includes tutorials with a bunch of examples, from simple to a little less simple, and some larger-scale applications. Sentiment computation, aggregation, diagnostic tools, visualization, regression – it’s all in there.

Readings

Check out the Research section, especially our vignette which explains the ins and outs of the software package along with accompanying code examples. The complete documentation can be found on the sentometrics CRAN page.

Shiny app

You might also want to have a look at the sentometrics.app package. Its sento_app() function embeds a Shiny application that displays many of sentometrics’ functionalities. Enjoy!

Media

Earlier versions of the package were presented as a lightning talk at the eRum 2018 (Budapest) and useR! 2019 (Toulouse) conferences, and recorded!