Multi-space matchmaking
Overview
Multi-space matchmaking uses additional input data that represent spaces and categories filters.
To enable Multi Space Matchmaking: Go to Admin Panel → Networking → Matchmaking → Enable Toggle Multi-Space Matchmaking:
Example of Set Up
In Multi Space we have included spaces/category filters for our matchmaking algorithm. In the case below, a space is a group of categories that you can find as a category filter. For example, “Country, Region and Appellation” is a space (a.k.a category filter) and “France” is the category of this space.
Example of categories filters
Example of the structure of category filter
Example of Matching
Inputs
In our example, we use 3 categories of filters/spaces. These are the product categories defined/grouped in filters/space:
Country, Region & Appellation
Grape Variety
Price range
So, we have the baseline user with the following interest categories.
Country, Region & Appellation
Italy
France
Grape Variety
Muscat
Pinot noir
Price range
Entry-level
Mid-range
Premium
Also, we have potential users for recommendation with the following activity categories.
User 1 → Full Match with baseline user | User 2 → Full Match per Spaces, but not complete with categories | User 3 → 2/3 match per spaces | User 4 → 1/3 match per spaces | User 5 → 0/3 match per spaces |
|---|---|---|---|---|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Expected results
The expected results indicate that User 1 is the strongest match, followed by User 2 and User 3 with decreasing match strengths. Users 4 and 5 have weaker or no matches in any of the spaces. The matchmaking algorithm considers both the number of matching spaces and the completeness of category matches within those spaces to determine the overall match strength.
Rank | User | Spaces Matches | Category Matches |
|---|---|---|---|
1 | User 1 | 3 | 7 |
2 | User 2 | 3 | 3 |
3 | User 3 | 2 | 4 |
4 | User 4 | 1 | 2 |
5 | User 5 | 0 | 0 |