A browser extension providing links to accurate information about newspaper health stories.
The NHS provides an really excellent service called 'Behind the Headlines'. They analyse scientific and health stories that have appeared in the press and give a critical assessment of whether the science supports the headlines.
Because the posts on Behind the Headlines link to both the original articles and to the science, it's possible to use them to build a database linking health stories and scientific papers. There are some 1300+ posts on the Behind the Headlines blog.
We wrote a browser extension that allows readers of news stories to see links to Behind the Headlines blog posts when one exists for the story in question.
We pulled the links to the science out of the relevant Behind the Headlines posts and showed them right next to the story itself.
An example showing an article on the Daily Mail website, with the NHS choices context next to it:
Also, using some simple keyword analysis of the article text we displayed graphs of relevant datasets taken from the NHS HES online service.
Project members: @beng, @grahamashton, @premasagar, @sshingler
Ideas for taking this project forward
Provide the APIs that we used to build these services as public APIs.
Package up the toolbar for distribution.
Spit and polish.
Update automatically when new posts appear on Behind the Headlines.
Cover more newspapers.
Suggest related Behind the Headlines posts based on keyword analysis of article texts.
Package as a widget.
Estimated costs for taking this project forward
Hard to say without knowing more about the NHS systems.
But this is a largely self-contained project to provide a toolbar and api using existing data. I'd imagine it would be no more than the other NHS choices widgets that have been developed.
About the data used for this project
I would have liked to use the NHS newsarticles API, but as usual, the api provides less data than simply scraping the website, so we did that instead.
This isn't really a criticism of the API - it was used to get hold of a list of articles to scrape and worked well. And likely would come in handy to turn the service into something more sustainable.
But, as important as an API misght be, clean webpages that are amenable to scraping are, in my opinion, equally important. Yours were pretty good in this regard.
I found the documentation of the API to be quite poorly written. I've done a lot of API integration and so on, and so could fill in the blanks myself. But for someone who was coming to this for the first time, I think it would be confusing. There's no shortage of developers - many at rewired state - who could help you to rewrite the documentation from the point of view of an integrator.
Finally, there's no reason on earth to require API keys or usernames and passwords to access data that's exactly the same data as is published on public websites. If you want to encourage people to build small, lightweight, experimental, one-off, hobby projects, you should remove these sources of friction. Right now.
Add photos by tagging them with rewiredstate:project="nhs-toolbar"