Matching patterns lemma and lower

, ,
1

I'm using LEMMA patterns but I'm noticing that it seems to be case sensitive, I would like to match on LEMMA and LOWER with a single pattern. Is this possible? Hoping it is to save myself a lot of time creating LOWER patterns of every variation.

Example:
{"pattern": [{"lemma": "aye"}], "label": "VOTE"}
will not match "AYE"

I have many different things that I'm trying to match that are either uppercase or lowercase that I would like to use the LEMMA on instead of LOWER due to their context in the sentence.

edit Maybe there is something I'm not understanding how LEMMA works because it seems to be matching some of my text regardless of case sensitivity but other text it will not match like the "AYE"

Thanks!

2

Hi! I think this might come down to a subtlety of how the lemmatizer works and whether it has a lemma available for the given token, or whether it defaults to the token text.

If you want to match on the lemma attribute but case-insensitive, you could try and use the REGEX operator and do something like this:

 {"pattern": [{"lemma": {"regex": "(?i)aye"}}], "label": "VOTE"}
1 Like
3

Thank you @ines!! I'm fairly new to all this so I appreciate your help!!