Need help in choosing the pipeline

Rasa Open Source
1

Hi All,

I am new to ML & AI / Rasa and trying to develop a company-policy specific faq chatbot.

My requirements are to develop a bot:-

  1. Learn and chat with company-specific policies faq.

  2. Learns and chats with small chitchat conversation.

I am not sure which policies and pipeline so i have used basic settings from weather bot or other tutorials.

 # Configuration for Rasa NLU.
 # https://rasa.com/docs/rasa/nlu/components/
 language: en
 pipeline: supervised_embeddings
 
 # Configuration for Rasa Core.
 # https://rasa.com/docs/rasa/core/policies/
    policies:
   - name: MemoizationPolicy
     max_history: 5
   - name: KerasPolicy
     epochs: 400
     batch_size: 100
     validation_split: 0.2
     max_history: 5
   - name: MappingPolicy

supervised_embeddings internally uses below component.

language: "en"

pipeline:
- name: "WhitespaceTokenizer"
- name: "RegexFeaturizer"
- name: "CRFEntityExtractor"
- name: "EntitySynonymMapper"
- name: "CountVectorsFeaturizer"

  1. How can i choose which component should be used and in which order’s.

  2. when i ask questions for chitchat using the keyword from faq intent NLU choose faq intent for it.

After reading lots of articles i am confused which policy to and pipeline to use.

2

Can any body please help

3

I would recommend to use default nlu pipeline and core config

4

Ghostvv - I am already using nlu pipeline .I wanted to know which is better supervised or pretrained for company specific bot which will have faq question and small talk chittalk.

5

it depends on your type of text you have. It is better to perform comparison tests to see which pipeline performs better