Simple but effective explainable test classication by taking the gradients of an embedding text input with respect to the output.
The higher the gradient of a single word, the more influence it had on the output; if the output probability is > 0.5, the rating is considered positive, and below that, negative.
Examples:
-
Notifications
You must be signed in to change notification settings - Fork 0
AloneTogetherY/explainable-text-classification
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
Simple but effective explainable text classication with tensorflow
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published