Spelling mistake create new pipeline

Hi ,i need create a Custom pipeline for spell correction, is there any implemented for example fuzzywuzzy python library or is necessary write a new Custom pipeline?. I listen opinions. other opcions is create new pipeline among all, since I notice that it is a general problem.

You can make one custom component with pyspellchecker, it has the advantage that you can add your words and create a personalized dictionary. The way I made this, it should be added at the beginning of your pipeline.

from rasa.nlu.components import Component
from rasa.nlu.training_data import Message
import typing
from typing import Any, Optional, Text, Dict

if typing.TYPE_CHECKING:
    from rasa.nlu.model import Metadata

from spellchecker import SpellChecker

class SpellCheckerEN(Component):

    provides = ["text"]
    defaults = {}
    language_list = ["en"]

    def __init__(self, component_config=None):
        super(SpellCheckerEN, self).__init__(component_config)

    def process(self, message, **kwargs):
        mt =  message.text
        str = mt.translate(mt.maketrans('', '', '!\"#$%&\'()*+,.-:;<=>?@[\]^_`{|}~'))
        words = str.split(' ')
        spell = SpellChecker(language=None)
        spell.word_frequency.load_dictionary('my_en_dict.gz')
        
        for word in words:
            if word not in spell:
                mt = mt.replace(word, spell.correction(word))

        message.text = mt

    @classmethod
    def load(
        cls,
        meta: Dict[Text, Any],
        model_dir: Optional[Text] = None,
        model_metadata: Optional["Metadata"] = None,
        cached_component: Optional["Component"] = None,
        **kwargs: Any
    ) -> "Component":
        if cached_component:
            return cached_component
        else:
            return cls(meta)

thank for the code

You are running a local installation of Rasa? I want to try this out, but we are running a Docker environment and I think I need to find a way to install the pyspellchecker module into the Rasa container… anybody tried this already?

Hey @wim.de.webmaster,

Actually i am building Docker environment and having trouble with it. would you mind sharing some of your material on that?