Entity extraction from text

Rasa Open Source
1

I want to know the maximum number of entity values that can be extracted from the Text using Rasa NLU model

2

Welcome to the forum! There’s no actual maximum on the number of entities or the values an entity can have. There are multiple entity extractors that have been integrated with Rasa. Each have their own characteristics depending on your use case. You can read about them here.

3

Thank you Greg Stephens. But I am getting only four values from the text .I have used CRF Entity Extractor . My OS is Ubuntu 18.04. I have used Rasa with Docker.

4

Please post your intent definition and the domain.yml fragment with the entity definitions.

5

language: “en”

pipeline:

  • name: “SpacyNLP”
  • name: “SpacyTokenizer”
  • name: “RegexFeaturizer”
  • name: “SpacyFeaturizer”
  • name: “CRFEntityExtractor”
  • name: “EntitySynonymMapper”
  • name: “SklearnIntentClassifier”
  • name: “DucklingHTTPExtractor” url: “http://duckling:8000”

The training data format I have used is json as per the documentation in the rasa official site.

6

The model of the dataset I have used is

{ “rasa_nlu_data”: { “common_examples”: [ { “text”: “EPFO site UPS not pinging. IP 10.174.11.164. test POE switch is down. Need to check the backhaul/power and poe switch.”, “intent”: “0”, “entities”: [ { “start”: 0, “end”: 13, “value”: “EPFO site UPS”, “entity”: “UPS” }, { “start”: 14, “end”: 25, “value”: “not pinging”, “entity”: “problem” }, { “start”: 27, “end”: 43, “value”: “IP 10.174.11.164”, “entity”: “IP” }, { “start”: 50, “end”: 60, “value”: “POE switch”, “entity”: “Switch” }, { “start”: 64, “end”: 68, “value”: “down”, “entity”: “problem” }, { “start”: 88, “end”: 102, “value”: “backhaul/power”, “entity”: “Switch” } ] } ] } }