Our analytical dataset focuses on collating quality news-related data. We have made it open-source so users can see our data and use it as a reference. Our dataset is open-source and freely available.
Check out our interactive search page/demo and/or codebook
Our JSON dataset is available for testing purposes only https://raw.githubusercontent.com/CuratedNews/analyticaldataset/main/CuratedNewsDataset.json
{
"title": "The three challenges keeping cars from being fully autonomous",
"link": "https://mittr-frontend-prod.herokuapp.com/s/613399/the-three-challenges-keeping-cars-from-being-fully-autonomous/",
"date": "2020-09-24 00:00:00 UTC",
word count, sentimentr score (whole text), & overall sentiment of article headline
"titlewordcount": 9,
"titlesentiment": "0.333333333333333",
"titlesentimentoverall": "Positive",
source, topic, & leaning
"Source": "MIT",
"Topic": "Technology",
"Leaning": "Academic",
"President": "Trump"
}
{"title": "The three challenges keeping cars from being fully autonomous","link": "https://mittr-frontend-prod.herokuapp.com/s/613399/the-three-challenges-keeping-cars-from-being-fully-autonomous/","date": "2020-09-24 00:00:00 UTC","titlewordcount": 9,"titlesentiment": "0.333333333333333","titlesentimentoverall": "Positive","Source": "MIT","Topic": "Technology","Leaning": "Academic","President": "Trump"}
- Overall positive = sentimentr score > 0
- Overall negative = sentimentr score < 0
- Overall neutral = sentimentr score = 0
- Check our interactive search page/demo for a hands-on with explanations of our analytical dataset
- We've made a headlines textual classifier with this dataset. Check out the demo.
Want to contribute? You can add unique categorical variables to the current data. Send us a pull request.
Visit https://curatednews.xyz