The links that you search for, must full urls of the form "https://en.wikipedia.org/wiki..."
I really like things that help people learn more and better, so this hack is really just an extensive of a larger life goal to help people learn. I love the thought that technology can empower someone to have a more interesting, better informed day.
How it works
Generating quizzes for any corpus of text is an excessively hard problem, but Wikipedia is a large, standardized, and classified collection of text sources which makes problem a lot easier to manage. So I pull in the article, scrape it clean, analyze the clean article text for specific patterns, and then generate questions from the information that I gather.
I altered all of the embedded Wikipedia links in the article content to redirect to WikiQuiz, so you can explore Wikipedia as a set of quizzes.
Challenges I ran into
- I ran into a lot of problems trying to parse out clean text from the article.
- I struggled then pulling salient questions from the data source.
- And displaying the content in a way that made sense.
Accomplishments that I'm proud of
I think that I have certainly made an excellent go of building a system that dynamically creates quizzes based on the received input.
What I learned
I learned a lot about navigating other people's messy html hierarchies.
What's next for WikiQuiz