This is the code of a simple Python Flask app that provides a demo of the m3inference Python library.
You probably are not interested in this repository, but instead should go to m3inference or read the WebConf (WWW) 2019 paper Demographic Inference and Representative Population Estimates from Multilingual Social Media Data that created m3.
Please visit a live copy of the web demo at http://www.euagendas.org/m3demo/
Clone this repository and then run
pip install -r requirements.txt
FLASK_APP=application.py python -m flask run
This demo requires Python>=3.6
Please cite our WWW 2019 paper if you use this package in your project.
@inproceedings{wang2019demographic,
title={Demographic Inference and Representative Population Estimates from Multilingual Social Media Data},
author={Wang, Zijian and Hale, Scott A. and Adelani, David and Grabowicz, Przemyslaw A. and Hartmann, Timo and Fl{\"o"}ck, Fabian and Jurgens, David},
booktitle={Proceedings of the 2019 World Wide Web Conference},
year={2019},
organization={ACM}
}
We use issues on this GitHub for all questions or suggestions. For specific inqueries, please contact us as [email protected]. Please note that we are unable to release or provide training data for this model due to existing terms of service.
This source code is licensed under the GNU Affero General Public License, which allows for non-commercial re-use of this software. For commercial inquiries, please contact us directly. Please see the LICENSE file in the root directory of this source tree for details.
This source code includes a copy of jQuery-Autocomplete. See the file or linked repository for further details about its licensing.
We also use Vizsla, a simple JavaScript API for Vega-Lite.
This web demo was created by Scott A. Hale and Graham McNeill. The research leading to M3 was carried out by Zijian Wang, Scott A. Hale, David Adelani, Przemyslaw A. Grabowicz, Timo Hartmann, Fabian Flöck, and David Jurgens. We thank the Volkswagen Foundation and the EPSRC for supporting the research.