You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Feb 20, 2023. It is now read-only.
Currently profiles are matched against full text, including name, age and place. However, there are several issues with that:
Place might change it's format (📍100м distance or exact name)
Age might increment over a long period of time
Text might be empty
Text might be non-unique, some examples of that are single emojies, common phrases, which leads to connecting profiles from different people
Text may change slightly over a long period of time: for example, user may correct grammar mistake and the app would never be able to connect new profile to older one
Solution
Maybe it makes sense to calculate image-specific type of hash on profile's photos and store them with profile's info. Then we can add matching weights, calculate text uniqueness and based on all these parameters make a decision whether this specific profile is connected to another one.
The text was updated successfully, but these errors were encountered:
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Summary
Currently profiles are matched against full text, including name, age and place. However, there are several issues with that:
📍100м
distance or exact name)Solution
Maybe it makes sense to calculate image-specific type of hash on profile's photos and store them with profile's info. Then we can add matching weights, calculate text uniqueness and based on all these parameters make a decision whether this specific profile is connected to another one.
The text was updated successfully, but these errors were encountered: