What would be the ideal configuration for a chatbot which deals with domain specific requests

Hi I have created a chatbot which performs various actions which are domain specific…I am trying to increase the intent recognition by editing the configuration…Can someone please help me out…

@nik202 Could you help me out?

@CAkhil can you share some code as a reference ?

@nik202 Hi…Could u tell the files which i should share so that I can share accordingly

@CAkhil as related to your issue? or steps you have followed.

@nik202 This is my current config file which I am trying to update to increase intent and entity recognition accuracy of my bot :

# Configuration for Rasa NLU.
# https://rasa.com/docs/rasa/nlu/components/
language: en

# # No configuration for the NLU pipeline was provided. The following default pipeline was used to train your model.
# # If you'd like to customize it, uncomment and adjust the pipeline.
# # See https://rasa.com/docs/rasa/tuning-your-model for more information.
  - name: WhitespaceTokenizer
  - name: RegexFeaturizer
  - name: LexicalSyntacticFeaturizer
  - name: CountVectorsFeaturizer
  - name: CountVectorsFeaturizer
    analyzer: char_wb
    min_ngram: 1
    max_ngram: 4
  - name: DIETClassifier
    epochs: 100
  - name: EntitySynonymMapper
  - name: RegexEntityExtractor
  - name: ResponseSelector
    epochs: 100
  - name: FallbackClassifier
    threshold: 0.65

# Configuration for Rasa Core.
# https://rasa.com/docs/rasa/core/policies/
   - name: MemoizationPolicy
   - name: TEDPolicy
     max_history: 5
     epochs: 100
   - name: RulePolicy
  #  - name: TwoStageFallbackPolicy
  #    nlu_threshold: 0.6
  #    core_threshold: 0.6
  #    fallback_core_action_name: "action_default_fallback" 
  #    fallback_nlu_action_name: "action_default_fallback" 
  #    deny_suggestion_intent_name: "out_of_scope"

I am trying to find the best possible configuration for my chatbot which deals with a domain specific actions

@CAkhil now elaborate your use case what you basically need whilst increasing the intent recognition and what is the significance?

@nik202 My chatbot has very similar intents which are differentiated by the entities given… Ex : Create a project - intent 1 Create a project with name testing1 - intent 2 create a project with name testing1 and add akhil@gmail.com - intent 3 because of which its getting confused

@CAkhil are you using forms for such use case?

@nik202 Yes…I am using forms but also there are very similar intents which have completely different functionality :

Example :
1)Add members to project 
   Add members to a VDI
2)Create project
   Create VDI

@CAkhil sorry for the late response, missed your message in different tags. Can you share some intents with examples?

@nik202 …Please find the following intent examples:

- intent 1: create_vdi_everything_given_except_type
  examples: |
    - create [vdi](vdi) in project [Create_Vdi](form_project_id) and add [Prathmesh.Tirodkar@lmtinfotech.com](form_member_list), [abhishek.som@lntinfotech.com](form_member_list) members
    - assign [vdi](vdi) in project [Test_Project_Collab](form_project_id) and add [Prathmesh.Tirodkar@gmail.com](form_member_list), [abhishek.som@lntinfotech.com](form_member_list) members
    - provision [vdi](vdi) in project [Demo_4march22](form_project_id) and add [Prathmesh.Tirodkar@outlook.com](form_member_list), [abhishek.som@lntinfotech.com](form_member_list) members

- intent 2: create_vdi_everything_given
  examples: |    
- create [personal](vdi_type) [vdi](vdi) in [project](project) [vdi_test](form_project_id) and add [Prathmesh.Tirodkar@lntinfotech.com](form_member_list), [abhishek.som@lntinfotech.com](form_member_list) members    \
- assign [pooled](vdi_type) [vdi](vdi) in [project](project) [Test_Project_Collab](form_project_id) and add [Prathmesh.Tirodkar@gmail.com](form_member_list), [abhishek.som](form_member_list) members    
- provision [personal](vdi_type) [vdi](vdi) in [project](project) [Test_Project_Collab](form_project_id) and add [Prathmesh.Tirodkar@outlook.com](form_member_list), [c.akhil@lntinfotech.com](form_member_list) members

@CAkhil thanks for the share, can you increase the examples to upto 10+ and train the model again whilst deleting older trained models?

@nik202 … I actually have more than 10 examples for each intent…but as many intents are very similar to each other (mostly vary by an entity value) , sometimes the bot is getting confused by entities and intents

@CAkhil try see the lookup table for other entities, I will see your issue may be over the weekend, as I never got such an issue on my projects. Do remind me please just whilst tagging me.

@nik202 any update ?

@CAkhil I’m really sorry Akhil, I am quite busy nowadays, maybe I will look this may be this weekend.