The sentimentIt package interacts with the SentimentIt API to analyze text and capture latent traits of interest. The SentimentIt system is designed to estimate a researcher-specified latent trait of interest for text corpora. The system is introduced in Carlson and Montgomery ''A pairwise comparison framework for fast, flexible, and reliable human coding of political texts.'' The system crowd-sources pairwise comparisons of the documents, currently through Amazon Mechanical Turk.
To install directly from GitHub, run the following code:
devtools::install_github('carlson9/SentimentIt/sentimentIt')
Alternatively, download the .tar.gz file from the main repository and install the package from source:
install.packages('sentimentIt_0.1.tar.gz', repos = NULL, type ='source')