Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Suggested Cargo's guesses at industry category aren't great. #79

Open
bowdidge opened this issue Feb 19, 2016 · 1 comment
Open

Suggested Cargo's guesses at industry category aren't great. #79

bowdidge opened this issue Feb 19, 2016 · 1 comment

Comments

@bowdidge
Copy link
Owner

Currently, the suggested cargo dialog will guess about the likely industry so it can suggest appropriate cargos. The current success rate for a training set is 60% of the guesses are correct, and 85% have the correct industry in the top three suggestions.

We should see if we can improve these success rates.

@bowdidge
Copy link
Owner Author

Currently, SwitchList uses a machine learning technique based on similar words to make the guess. (There's also a couple heuristics that strengthen certain terms.)

I tried switching to an approach that looks for similar character pairs; this turned out to be much worse (success rate of 30% on same training set) because cases like "paving" and "canning" appeared to group together because of the "ing" similarity. Other approaches (similar strings) fail in similar ways.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant