Entity Extraction in Rasa

Question

I want to create a chatbot that extract words from the user message as entity and send it to dictionary and in return get the meaning of that word.

But the problem is entity values are not getting extracted, and I am getting empty brackets [ ]. I am trying to solve this issue for weeks now. Now, I am exhausted and desperate. Please help me to figure this out.

Here is all the files: https://github.com/Attiqakaleem0/Rasa-word-meaning-bot

Installation versions on my system are: Rasa Version : 3.6.13 Minimum Compatible Version: 3.5.0 Rasa SDK Version : 3.6.2 Python Version : 3.10.0 Operating System : Windows-10-10.0.19045-SP0

Answer 1

You need to add the RegexEntityExtractor to the pipeline in your config.yml.

config.yml:

pipeline:
- name: RegexEntityExtractor
  case_sensitive: False
  use_lookup_tables: True
  use_regexes: True
  use_word_boundaries: True

Additionally you will need to modify your NLU training data to match the correct format for extracting entities.

Before:

Clarify the term [sympathy] for me. (term)

After:

Clarify the term [sympathy](term) for me. 

The Rasa documentation goes into more detail about how to format NLU training data.