Creating user recommendation system

Hello all,

I am trying to create a recommendation system of users to a group and of groups to groups. For users, they can have attributes of A, B, and C for example whereas groups can have attributes of 1, 2, and 3.

I would like to generate a list of recommended users to a group to possibly add this user. For example, I would want a group with an attribute of 1 to be recommended users with attributes of A and B. However, if the group has an attribute of 2, perhaps it gets recommended only users with attributes of B and C. The importance of this recommendation system is that it is not always 1 matches to 1 or 1 matches to A.

I have been trying for hours to do this in a list with “AND” and "OR"s. However, I can never seem to get the data linked for both users and groups to allow me to specify which attributes for the respective user and group needs to match up to display the recommended users.

Think of creating a recommendation system like how Facebook recommends relevant friends. Is this even possible in Adalo?

Thanks all!!

Hi @Christopher99 ,

You might want to check this out first,

I’ve been trying to wrap my head around this. It does get me closer though (I think).

To make it more simple, I just trying to filter a simple list that displays content of a collection based on a mutual relationship that exists in another database.


Collection 1 (The one I want displayed) has a field with a M:M relationship to Collection 3.

Collection 2 (what determines the filtering) has a field with a M:M relationship to Collection 3.

Collection 3 (the mutual field) has fields with a M:M relationship to both. It also has a M:M relationship to itself.

I want to display a filtered Collection 1 if Collection 1 has the same field content in Collection 3’s M:M self-relationship AND it contains one of Collection 2’s fields with a M:M relationship to Collection 3.

It is better to come up with examples.

The advantage is we can use it for testing, if the result is the same or not.

We need to be as specific as possible rather than metaphoric.