Fix regex matching being used in PartOfSpeech representation model

[woranov]

What does this PR do?

By default, pandas' str.contains interprets the argument as a regex pattern. This causes the PartOfSpeech model to error out if the topics contain e.g. mismatched parentheses. This PR fixes that issue.

Fixes #2153.

Before submitting

• This PR fixes a typo or improves the docs (if yes, ignore all other checks!).
• Did you read the contributor guideline?
• Was this discussed/approved via a Github issue? Please add a link to it if that's the case.
• Did you make sure to update the documentation with your changes (if applicable)?
• Did you write any new necessary tests?

[MaartenGr]

Thank you for the PR. I didn't see an open issue attached which is generally the procedure when opening up a PR. Could you add it?

With respect to the suggested change, do you have any example of what it fixes? I'm not sure I understand in wich specific scenarios the issue would happen.

[woranov]

Apologies! Indeed there are some preconditions to be met for the error to arise.

Opened an issue with a reproduction example in #2153.

[MaartenGr]

Thanks for the update! This seems good to me, let's merge 😄