This project presents the RedDust data resource consisting of personal attribute labels for over 300k Reddit users across five predicates: profession, hobby, family status, age, and gender.
We construct RedDust using a diverse set of high-precision patterns and demonstrate its use as a resource for developing learning models to deal with implicit assertions.
RedDust consists of Reddit users ids, the corresponding users' personal attribute labels, and the users' post ids, which may be used to retrieve the posts from a publicly available crawl or from the Reddit API.
The link to the dataset itself: https://zenodo.org/record/3234005#.XO2Cm6WxXRZ
In this repo we put the accompanying code used to extract RedDust data via Patterns
Use hadoop to run the scripts For each attribute use
-mapper mapper_<att_name>.py - reducer joint_reducer.py
Reducer is the same for all. For further details about the patterns refer to the pdf of the paper
Then use code from
https://github.com/Anna146/HiddenAttributeModels/tree/master/prepare_data/hadoop
to fetch the bodies of the messages instead of their ids.