sentometrics: An Integrated Framework for Textual Sentiment Time Series Aggregation and Prediction

Introduction

The sentometrics package is an integrated framework for textual sentiment time series aggregation and prediction. It accounts for the intrinsic challenge that textual sentiment can be computed in many different ways, as well as the large number of possibilities to pool sentiment into a time series index. The package integrates the fast quantification of sentiment from texts, the aggregation into different sentiment time series, and the prediction based on these measures. All in one coherent workflow!

See the package website and the vignette published in the Journal of Statistical Software for plenty of examples and details. We also refer to our survey organized as an overview of the required steps in a typical econometric analysis of sentiment from alternative (such as textual) data, and following companion web page.

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("SentometricsResearch/sentometrics")

Shiny application

For a visual interface as a Shiny application of the package's core functionalities, install the sentometrics.app package, and run the sento_app() function.

Reference

Please cite sentometrics in publications. Use citation("sentometrics").

Acknowledgements

This software package originates from a Google Summer of Code 2017 project, was further developed during a follow-up Google Summer of Code 2019 project, and benefited generally from financial support by Innoviris, IVADO, swissuniversities, and the Swiss National Science Foundation (grants #179281 and #191730).