Current version: v1.0 (released in XX month XX year)
Terms of use: Uses are subject to Reddit API terms.
Click here to download v1.0 (XX GB)
This file contains JSON files containing
… along with XX license and Reddit API terms of use.
- README.txt
- TERMSOFUSE.txt
- metadata/
├── subreddit 1/
├── <reddit.post.id 1>.json
├── <reddit.post.id 2>.json
├── ...
└── ...
├── subreddit 2/
├── ...
└── ...
Each <reddit.post.id>.json contains metadata of image posts from a single subreddit,
following this schema.
A sample example from the datatset:
{
"title": "elon musk at my local barnes and noble",
"url": "https://i.redd.it/a7oueb592bpb1.jpg",
"comments": [
{
"author": "moxyfloxacin",
"body": "i hope its pages and pages of the truth i doubt it though",
"created_utc": "2023-09-19 20:46:15",
"analysis": {
"common nouns": {
"hope": 1,
"page": 2,
"doubt": 1,
"truth": 1
},
"emotion tag": "optimism",
"attention scores": {
"hope": 1.0,
"doubt": 0.11211709678173065
}
}
},
{
"author": "DataAstronaut_",
"body": "just listened to walter isacsons podcast with lex friedman from the sound of him talking about elon on it should be an interesting read",
"created_utc": "2023-09-19 21:22:38",
"analysis": {
"common nouns": {
"walter": 1,
"isacsons": 1,
"friedman": 1,
"elon": 1,
"podcast": 1,
"walter isacsons": 1,
"interesting read": 1,
"walter isacsons podcast": 1,
"isacsons podcast": 1
},
"verbs": {
"listen": 1,
"sound": 1,
"talk": 1,
},
"adjectives": {
"interesting": 1,
"lex friedman": 1
},
"emotion tag": "approval",
"attention scores": {
"listened": 1.0,
"read": 0.2401873618364334,
"interesting": 0.10860899090766907,
}
}
},
...
]
}