There exists an overwhelming amount of scientific evidence that makes discerning the truth very difficult. At times many different scientific articles may contradict. Furthermore, scientific evidence can be very hard to understand if you have no prior scientific research background, and is often misinterpreted. As science students, we want to bridge the scientific knowledge gap to make scientific research more accessible and interpretable. Thus, we created Rapid Evidence to address this challenge.
What it does
Rapid Evidence is a website designed specifically for the public. Individuals who are looking for scientific knowledge about a certain topic, treatment, disease will be able to enter a search query. That search query will then be entered into a web scraper that looks through Google Scholar, filters out articles to only include well cited and/or recent Systematic Reviews, and extracts results to be put into an understandable summary. To ensure that only the BEST evidence is included (best according to Cochrane's gold standard), only systematic reviews and meta-analysis with a GRADE of B+ or higher are included.
How we built it
We used Selenium for web scraping of Google Scholar. We also used photoshop to produce a preliminary website design and built a basic framework using HTML and CSS for the website.
Challenges we ran into
We spent the first 5 hours of the hackathon simply trying to install Selenium.
Accomplishments that we're proud of
Most of our team members were not familiar with Python and we surprised ourselves in our capability to grasp a basic understanding and had the ability to contextual examples online to our unique problem. From the process, we have learned to prioritize and time manage appropriately to get our project. Despite our fear of our limited abilities, we are very proud to have participated in this event and to have gone above and beyond in learning new skill sets in an applicable manner.
What we learned
We learned how to make a basic web-scraper. We also learned HTML and CSS for the first time!
What's next for Rapid Evidence
Currently, we have all the individual pieces of Rapid Evidence, but they do not fully run together. We plan on creating the website design and connect it with the backend web-scraper in order to make the website fully functional!