The sentometrics package is designed to do time series analysis based on textual sentiment. It accounts for the intrinsic challenge that, for a given text, sentiment can be computed in many ways, as well as the large number of possibilities to pool sentiment across text and time. This additional layer of manipulation does not exist in standard time series analysis and text mining packages. As a final outcome, this package provides an automated means to econometrically model the impact of sentiment in texts on a given variable, by first computing a wide range of textual sentiment time series and then selecting those that are most informative. The package created therefore integrates the qualification of sentiment from texts, the aggregation into different sentiment measures and the optimized prediction based on these measures.
Sentiment computation and aggregation into sentiment measures: sento_corpus, ctr_agg,
compute_sentiment, sento_measures, to_global
Sparse modelling: ctr_model, sento_model
Prediction and post-modelling analysis: predict.sentomodel, retrieve_attributions,
perform_MCS
The development version of the package is available at https://github.com/sborms/sentometrics.
Useful links: