LAST PROJECT

What New York Times comments over a week in August tell us about 2020 Dem candidates

ai2html assignment, Storytelling Week 2

By KASTURI PANANJADY

Published September 9, 2019

How New York Times Commenters Feel about 2020 Democratic Candidates

An analysis of text from comments on opinions pieces on the Times website between August 3-10.

Biden

0.6

Dem

Warren

Other

Yang

Size scales

with number

of comments

considered

Increasing

positive

sentiment

Harris

Buttigieg

Sanders

de Blasio

0.1

Klobuchar

0.7

0.3

Increasing negative sentiment

Candidates on

this line are as liked

as they are disliked

Source: NYT Community API

How New York Times Commenters Feel About 2020 Democrats

Sentiment analysis on Op-Ed comments from August 3-10.

Biden

0.6

Dem

Positive

Warren

Other

Yang

Size scales

with number

of comments

considered

Harris

Buttigieg

Sanders

de Blasio

0.1

Klobuchar

0.7

0.3

Source:

NYT API

Negative

Neutral

NYT Commenters and 2020 Democrats

Sentiment analysis on Op-Ed comments from August 3-10.

Biden

0.6

Dem

Warren

Other

Yang

Positive

Size scales

with number

of comments

considered

Harris

Buttigieg

Sanders

de Blasio

0.1

Klobuchar

0.7

0.3

Source:

NYT API

Negative

Neutral

Methodology: For this project, I scraped web urls for opinion pieces using Selenium, and then queried the Community API at the New York Times. I think isolated proper nouns from the comments and used regex to account for spelling mistakes. Finally, I ran the text through a sentiment analysis library and used that to classify comments.

On the X and Y axes is the harmonic mean of the negative and positive frequencies of scores for comments about Democratic candidates.