CS-Notes-byCKL
Primary version
Primary version
  • First Page
  • PHP Notes
    • How a Theme Is build
  • Applied Machine Vision
    • The Pinhole Camera Model
    • Linear Algebra Recap
  • NLU/NLP Chatbot powerd by Rasa for Telekom
    • RASA
      • Terminology and Definitions and structure of a bot
      • Rasa Core Vs Rasa NLU
      • Rasa NLU
        • Training and Data Format RASA NLU
        • Entity Extraction
        • Evaluation
      • Preprocessing Methods
        • Bag of Words
      • Setup
  • Evalutation/Testing
  • Distributed Sytems
    • Week 1
    • Java I/O
  • Diskrete Strukturen
    • 1. Kombinatorik
      • Kombinatorische Beweisprinzipien
      • Ziehen von Elementen aus einer Menge
      • Wichtige Zählprobleme
      • Rekursionsgleichungen
    • 2. Zahlentheorie
      • Teilbarkeit und Primzahlen
      • Modulare Arithmetik
      • Vermischtes
    • MIT Week 1 Proofs
    • 3. Graphentheorie
      • Modellieren von Graphen
      • Grundbegriffe
      • Bäume und Wälder
      • Graphen Eigenschaften
    • 4. Algebraische Strukturen
      • Verbände
      • Isomorphe und homomorphe Abbildungen
  • Fingerprint extraction of electrical appliances
    • 1. Data Preprocessing
    • 2. Transition Detection
  • AI in a nutshell
Powered by GitBook
On this page
  1. NLU/NLP Chatbot powerd by Rasa for Telekom
  2. RASA
  3. Rasa NLU

Training and Data Format RASA NLU

How does the structure for data Rasa NLU learns from look like?

PreviousRasa NLUNextEntity Extraction

Last updated 7 years ago

  • common_examples : used to train entity and intent models

  • regex_features : a tool to help the classifier detect entities or intents

  • entitiy_synonyms:

There are 3 ways to train a model

  • With a script

  • With own python code

  • With Http server

Overview of training data structure
Train data is used to teach the model how inents and entities look like.
Regular Expression can be used to detect common patterns or structures in the data like zip codes or urls.
Examples for Markdown Format
Traing a model with a script and its options