just discovered this forum and thought that I might ask my question here.
I’m struggling a bit with finding a technical solution for a matching patching for my web app and thought that you might help me out.
I’m creating a web app that matches like-minded people based on the info they share with the app. To put it simply, they describe themselves with hashtags. The most important data to match would be the user’s input, not names or their location.
Now I started to play around with some workflows, but am still stuck. Is it possible to develop a matching system with Bubble at the moment? Do I have to work with lists? There was someone who asked a similar question a while ago, but didn’t get a response.
Thank you so much. It means a lot to me because I’m a non-technical founder and try to at least build the prototype for my idea by myself.
Best regards from lovely Las Palmas,
Absolutely. You just need to get creative with workflows and data types. Keep at it and you’ll continue to get better!
Good to know, thanks. For example, I have three users with the following inputs:
User 1: A, B, C, D, E
User 2: E, F, A, G, H
User 3: I, J, B, C, E
Therefore, User 1 and User 3 would be the best match out of these three:
User 1 and User 3: 60%
User 1 and User 2: 40%
User 2 and User 3: 20%
How could I use Bubble in the most efficient way for this matching pattern? I really thought about creating lists for each user and then somehow comparing them to each other. Do you know of a better way to execute it?
Getting those kinds of percentage would require a compare on two users. Each list of hashtags would need to be combined with an added (A,B tag) then sorted and counted. But the real problem is if you wanted to see which other users you matched up best with. You’d need to do a compare on each. How many in your database? 1000 or 100,000. This would be end up being a VERY CPU intensive operation and ultimately a failed (impossible solution),
I suggest you try thinking of the problem differently. Start simply: with a list of allowable tags. Then later you can allow users to create new tags on the fly. Create a Thing called TagList where you keep count of all the Tags. For instance, every time someone added the tag, #drinksCoffee, you’d add one to the drinksCoffee count. And of course if the deleted the tag, you’d subtract one from the count.
Now you can present a list of all tags and their popularity to your users. Also, if you have a lot of tags, it would be good to also add a tag category.
Finally, you would need a way to search for some combination of tags. For instance: #drinksCoffee, #likesPhotography, #playsTennis. I can’t say for sure, but I think that kind of search would be much more doable with Bubble. Can’t say exactly how, as I have not gotten that far.
Maybe someone with more experience can chime in and say whether or not that kind of search is even possible. If not… time to rethink the solution. I’m sure there’s some way to provide a matching service.
Hi Townsend, sorry for the late response.
That sounds really good, I’ll think about that. Moreover, I wanted to create a list of tags beforehand in the back-end and allow users to search for these items with a Search Box. If their input is in the database, they can choose it and the specific list of that tag gets a +1 count. If a tag doesn’t exist, they can request it or better create it by themselves. Later on, you’d be able to edit your list of tags and perhaps delete some, too. Therefore, this would subtract -1 from the tag list.
I guess that this should be definitely doable with Bubble. I’ll try that out later!
@chrschw97, Were you able to make it work? I just started using Bubble to build a very niche product that uses a similar matching process as a core feature. Thanks!
Need help, the same problem here,