I'm trying to do research about arabic in rasa nlu with tensorflow

Rasa Open Source
1

data.md (1004 Bytes) config.yml (389 Bytes)

That’s my config and the data also, i’m using rasa nlu 0.15.1 I’m trying to research about arabic nlu with tensorflow pipeline but the result did not give the entity extraction,

04

anyone maybe have an idea

2

Hey @kunci115, welcome to the forum :slight_smile:

Before trying anything else, could you try correcting your config file? A line starting with “- name” is the name of a component, and the lines that follow below it are the options passed to the component. Your config could look like this (for clarity, I grouped things that belong together):

language: ar

pipeline:
  - name: CountVectorsFeaturizer
  
  - name: EmbeddingIntentClassifier
    intent_split_symbol: +
    intent_tokenization_flag: true


policies:
  - name: KerasPolicy
    batch_size: 50
    epochs: 200
    max_training_samples: 300

  - name: FallbackPolicy
    fallback_action_name: action_default_fallback
  
  - name: MemoizationPolicy
    max_history: 5
  
  - name: FormPolicy
  
  - name: MappingPolicy

Let me know if this helps.

2 Likes
3

it helps man! thanko you somuch!

4

@kunci115 Are you able to build Rasa for Arabic language? Is entity extraction working for Arabic?

5

yes its totally working bro @alswamy

6

@kunci115 could you share a sample arabic chatbot for a reference. that will be very helpfull. Thanks

7

Dear @kunci115 Plz share the sample bot for arabic. It will be very helpfull

8

i tried this but i got this error

ComponentNotFoundException: Failed to load the component ‘EmbeddingIntentClassifier’. Cannot find class ‘EmbeddingIntentClassifier’ in global namespace. Please check that there is no typo in the class name and that you have imported the class into the global namespace. Either your pipeline configuration contains an error or the module you are trying to import is broken (e.g. the module is trying to import a package that is not installed). Traceback (most recent call last): File “/Users/walaaalrefaee/Documents/GitHub/helpdesk-assistant/venv/lib/python3.7/site-packages/rasa/nlu/registry.py”, line 121, in get_component_class return rasa.shared.utils.common.class_from_module_path(component_name) File “/Users/walaaalrefaee/Documents/GitHub/helpdesk-assistant/venv/lib/python3.7/site-packages/rasa/shared/utils/common.py”, line 32, in class_from_module_path raise ImportError(f"Cannot retrieve class from path {module_path}.") ImportError: Cannot retrieve class from path EmbeddingIntentClassifier.

9

@walaareafaee the classifier has been deprecated, the helpdesk assistant now also uses the newer DIETClassifier, see the config.

1 Like