I am trying to implement Multi Label Classification on Australian Skill Classification dataset where I have to predict multiple job titles that fits the profile out of 600 titles for a single entity.
The issue is I am not able to convert "ANZSCO_Title" in a way that I can apply Multi-Label Classification on it.
Please suggest a solution.
I tried using LabelBinarizer directly on it but that was giving me 598 zeroes and 1 one per row.
I need a y-label with these 600 titles in seperate columns being marked 1 or 0 based on if it fits the score or not.
For reference currently "ANZSCO_Title" looks like this: Dataset Sample
But I want all the titles to be seperate columns so that I can apply MultiLabelBinarizer on it.