Sentiment Textual data - useR! 2019user2019.r-project.org/static/pres/lt255084.pdf · 2019-07-26 · Textual data Sentiment. The R package sentometrics to analyze textual sentiment

Textual dataSentiment

The R package sentometricsto analyze textual sentiment

AggregateCompute Predict

Sentometrics is about econometrics meeting sentiment!

Transformation of qualitative data into quantitative

sentiment variables

Application in econometric analysis of the relationships

between sentiment and other variables

Corpus

Word lists (lexicons)

Valence shifters

Within-document weighting

Sentiment

matrix

Metadata features

Sentiment

matrix

Frequency

Across-document weighting

Across-time weightings

Sentiment

time series

Further

aggregation

Many time series obtained across

lexicons, features and time aggregation schemes

lexiconfeature time

aggregation

target other variables sentiment

Model

Sentiment time series

Other variables

Model specifications

Output

Forecasting

Precision

Attribution

Insights

Corpus

Valence shifters

Sentiment matrix

Lexicons

Weightings

Frequency

Sentiment

time series

Specifications

Target variable

Output

Attribution

Use the sentometrics R package!

To aggregate textual data into

sentiment variables…

… & insights

Ardia, Bluteau, Borms and Boudt (2018). “The R Package sentometrics to Compute, Aggregate and Predict with Textual Sentiment”. Available at: http://dx.doi.org/10.2139/ssrn.3067734.

Repository: https://github.com/sborms/sentometrics.

Project website: https://www.sentometrics.org.

Samuel [email protected]