NYT Headlines Sentiment

Here's another example using the textblob library to look at NYT headlines.

The textblob library outputs two metrics:

Polarity

• Polarity measures the sentiment's direction on a scale from -1 to 1.

  • -1 indicates a very negative sentiment.

  • 0 indicates a neutral sentiment.

  • 1 indicates a very positive sentiment.

• This score is determined by analyzing the text's adjectives and other sentiment-bearing words. A text with a high number of negatively connotated words will have a negative polarity, while one with many positive words will have a positive polarity. Neutral statements, often factual or lacking strong emotional words, will have a polarity close to 0.

Subjectivity

• Subjectivity measures how much of the text is based on opinion versus factual information, on a scale from 0 to 1.

  • 0 indicates very objective text (fact-based).

  • 1 indicates very subjective text (opinion-based).

• This score is assessed by identifying phrases and expressions that reflect personal opinions or feelings, as opposed to statements of fact. The more a text relies on opinions, the higher its subjectivity score.

Interpretation of Polarity=0.0, Subjectivity=0.0

When a sentiment analysis returns Polarity=0.0 and Subjectivity=0.0, it suggests that the analyzed text is perceived as neutral and objective. This can happen when the text:

• Lacks emotionally charged words or phrases, leading to a neutral polarity.

• Contains mostly factual information without explicit opinions, biases, or subjective statements, resulting in a low subjectivity score.

To try this example, you'll need the following:

An API key from the NYT developer site. That's covered here

Install the textblob and requests library in the Terminal on your  Raspberry Pi.

$ pip3 install requests textblob

Install the sentiment analysis lexicon:

Using Thonny, create a Python file to download the sentiment lexicon. You only need two lines of code:

import nltk

nltk.download('punkt')

Save the code as something like nyt_lexicon_download.py and run it. It takes only about a second.

Add your API Key to the NYT sentiment analysis example and run the code.