Examples
Examples
Try the starting example
- JSON Eligible catalog item IDs
- ["A","B","C","D"]
- JSON Recommendations [user ID,item ID]
- [["U1","A"],["U1","B"],["U2","A"]]
- JSON Eligible user IDs
- ["U1","U2","U3"]
- Catalog coverage fraction
- 0.5
- User coverage fraction
- 0.666667
- Largest item exposure share
- 0.666667
- Out-of-catalog recommendation rows
- []
Try another scenario
- JSON Eligible catalog item IDs
- ["A","B","C","D"]
- JSON Recommendations [user ID,item ID]
- [["U1","A"],["U1","B"],["U2","C"],["U3","D"]]
- JSON Eligible user IDs
- ["U1","U2","U3"]
- Catalog coverage fraction
- 1
- User coverage fraction
- 1
- Largest item exposure share
- 0.25
- Out-of-catalog recommendation rows
- []
Calculation method
Calculation method
Catalog coverage fraction
len({r[1] for r in b if r[1] in a})/len(a)
User coverage fraction
len({r[0] for r in b if r[0] in c})/len(c)
Largest item exposure share
max(sum(r[1]==k for r in b) for k in a)/len(b) if b else 0
Out-of-catalog recommendation rows
J([r for r in b if r[1] not in a])
Symbols and units
- a
- JSON Eligible catalog item IDs
- b
- JSON Recommendations [user ID,item ID]
- c
- JSON Eligible user IDs
Use the units shown. A list keeps the order in which you enter its values. The symbol ^ means “raised to the power of”.
Calculation steps and notation
sum(list) adds all entries in the list.
Assumptions & limits
Assumptions & limits
- Exact item/user coverage of supplied recommendation exposures. Catalog and user denominators nonempty; repeated exposures count toward concentration. Supported limits: All numeric inputs must be finite. Only the explicitly described mathematical scenario is supported.
Are these inputs private?
These calculations run in your browser. A shared link contains the values you choose to share. Only questions submitted to the separate AI finder are sent to OpenAI.