Add regex notations directly to examples in lookup tables

Hello is there a way to add regex notation to lookup tables? For example ^gin$, currently if the users message has the characters “gin” such as in managing or aging, the entity gets trigger.

I am only interested in extracting keywords from users messages which is why I only have the RegexEntityExtractor in my pipeline. I would like to keep use_word_boundaries=False.


  • name: “RegexEntityExtractor” case_sensitive: False use_word_boundaries: False use_regexes: True

I figured it out. Changed lookup tables to regex. If anybody has a similar issue where they just want keyword matching. The config file uses the pipeline

name: “RegexEntityExtractor” -case_sensitive: False -use_word_boundaries: False -use_regexes: True

Then in your NLU file it’s:

Regex: trigger_keywords Examples: |

  • /bgin/b


Now if someone writes “I am managing this store” the entity is not triggered unless it has the word “gin” exactly.