Pilot studies for NHSx on understanding the digital shift and trends in End Of Life care due to the lockdown during the COVID-19 pandemic.
a) Automated filtering of news articles discussing remote and digital services, relating to the NHS
b) An analysis of Google Play Store reviews relating to GP (and similar) service apps
c) Automated filtering of news articles discussing End Of Life (EOL) care in the UK.
More information on studies a) and b) can be found in the following slide deck (Nesta staff only) In general, refer to these slides for methodological notes.
We use the NewsAPI API (note: this isn't free) to collect relevant news articles.
Please see __name__ == '__main__' in utils.keyword_filter for how to do this.
News articles are collected according to various querying strategies, as discussed in the methodological slide deck.
The following instructions assume a UNIX-like system (e.g. this works on both my MacBook and cloud Linux server). On Windows I would first recommend using the Windows-Linux Subsystem, and then using the Ubuntu interface on there.
Firstly, install conda on your local machine.
After this you can create an environment to work in:
conda create -y -n nhsx python=3.7 pandas
conda activate nhsx
pip install inflect
pip install newsapi-python
pip install openpyxl
Next, get the code from this repository:
git clone https://github.com/nestauk/nhsx_digital_shift
cd nhsx_digital_shift
mkdir secrets
export PYTHONPATH=$PWD
Now you will need a NewsAPI key. You should store this in a file called news_api_key and the contents of the file should be the key only. Now you should copy the file news_api_key which you created into the folder secrets, e.g.:
cp /path/to/news_api_key secrets
Finally, you can run the data collection and processing code:
python utils/keyword_filter.py
The output files will appear in your current working directory.