usnews: Texts (not) relevant to the U.S. economy

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

Description

A collection of texts annotated by humans in terms of relevance to the U.S. economy or not. The texts come from two major journals in the U.S. (The Wall Street Journal and The Washington Post) and cover 4145 documents between 1995 and 2014. It contains following information:

• id. A character ID identifier.

• date. Date as "yyyy-mm-dd".

• texts. Texts in character format.

• wsj. Equals 1 if the article comes from The Wall Street Journal.

• wapo. Equals 1 if the article comes from The Washington Post (complementary to ‘wsj’).

• economy. Equals 1 if the article is relevant to the U.S. economy.

• noneconomy. Equals 1 if the article is not relevant to the U.S. economy (complementary to ‘economy’).

Usage

data("usnews")

Format

A data.frame, formatted as required to be an input for sento_corpus.

Source

Economic News Article Tone and Relevance. Retrieved November 1, 2017.

Examples

data("usnews", package = "sentometrics")
usnews[3192, "texts"]
usnews[1:5, c("id", "date", "texts")]

sentometrics documentation built on Aug. 18, 2021, 9:06 a.m.