Returns an object of class c("tbl_df", "tbl", "data.frame")
containing one row for every sentence in the corpus.
The returned data frame includes at a minimum following columns:
"id" - integer. Id of the source document.
"sid" - integer. Sentence id.
The coreNLP backend also currently returns a column "sentiment" that
gives a score from 0 (worst) to 4 (best) for how positive the
tone of the sentence is predicted to be.
References
Manning, Christopher D., Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J. Bethard, and
David McClosky. 2014. http://nlp.stanford.edu/pubs/StanfordCoreNlp2014.pdf.
In: Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pp. 55-60.
Socher, Richard, et al. "Recursive deep models for semantic compositionality over a sentiment treebank." Proceedings of the conference on
empirical methods in natural language processing (EMNLP). Vol. 1631. 2013.
# how do the predicted sentiment scores change across the years?get_sentence(obama) %>%
group_by(id) %>%
summarize(mean(sentiment), se = sd(sentiment) / sqrt(n()))