I had an idea in mind for a while now. I would like to create a “hub”, a repository of science-backed studies on a specific topic. If you’re familiar with this website, examine.com does exactly what I intend to do on a broader scale.
While examine is not specific to one “niche” but rather gathered studies on specific supplements, I intend to do the same for diabetes and everything related to insulin resistance.
Ideally, the project would contain a Home page with a search bar and illustrations all around, with part of the Homepage linking directly to subcategories like:
studies repository for the tag « Cinnamon »
studies repository for the tag « Metformin »
studies repository for the tag « chromium »
I imagine I’ll have to create a database regrouping every study about the subject;
Ideally, I’d crawl/scrape the URLs of the major databases publishing studies: Pubmed, ScienceDirect, Google scholar/NCBI, and the likes.
I’m also looking for guidance regarding the best practices here in building this database if any of you have experience in similar projects.
For the other pages of the project, for each of the tag/keywords like those mentioned above, say « Metformin », guidance regarding the posology as well as a data frame summarising the outcome of each study performed on Metformin would be enough for starters.
Ideally, each row of the data frames would contain these studies’ output/conclusion, with columns returning information like « limits of the study (done in rats and not in human, for instance) », « notes », etc. This would be enough, to begin with.
All in all, I’d like to just build an « MVP » for now, with basic features. Ultimately, advanced data would be attached to these studies and I would take care of this as I have a background in data science/analysis.
Would airtable be the ideal tool for the backend part?
I’ve heard there may be limits with the ability to requests data from airtable (a few requests per second)
Would that be an issue if too many users were to try to access the databases of compiled studies about a specific keyword?
Any tips appreciated :slightly_smiling_face: