Weather text classifier with machine learning and natural language processing
2.8K
https://github.com/jramcast/nuublo-predict
Weather text classifier. From a tweet text, predicts the kind of weather event, the sentiment, and whether the text talks about past present or future.
docker run -d -p 8752:8752 jramcast/nuublo-predict
curl -X POST \
-F text="Partly sunny, breezy with highs in the 80s. Heavy and possibly severe t-storms around and after sunset. Watches could be issued later." \
http://127.0.0.1:8752/prediction
Result (detected classes):
[
"WEATHER: sun",
"WEATHER: hot",
"WEATHER: storms",
"WEATHER: wind",
"WHEN: future (forecast)",
"WHEN: current (same day) weather",
"SENTIMENT: neutral / author is just sharing information"
]
Content type
Image
Digest
Size
564 MB
Last updated
about 9 years ago
docker pull jramcast/nuublo-predict