see more tags like this...:
[malayalam-tamil-epubs-good]
Then this passage will get some tags like this:
Until you find some passage which gets tags like this:
[malayalam-tamil-epubs-bad]
Also, the tags you see are removed and can be re-applied and sorted in any order.
A:
Tags are selected/sorted by domain. Your tag example is referring to the domain of the tags, so it is sorted by that. The top tag is the one which most matches.
The only difference between good and bad is the number of applications. If it is the first time you are seeing that one, then it is not applied until you fix the problem that needs to be applied to achieve a good tag.
The tagger is using that domain to determine which tag to show. To see the main tag, set the domain to a empty string:
domain = ''
However, this will make the tagger pick every tag with the same domain as the current one. You may want to set a different domain, like the book title:
domain ='malayalam-tamil-epubs-good'
Now you have an application of the tagger to show the tag you want.
apiVersion: machinelearning.seldon.io/v1
kind: SeldonDeployment
metadata:
name:'seldon-classification-demo'
spec:
componentSpecs:
- label: Classification
component: classification-component.seldon.io
parameters:
- in: 'body'
name: body
required: true
schema:
properties:
input:
description: >
Related links:
Comments