Can I add more data to snips-nlu on the fly? - python-3.x

Question

I am using snips-nlu to create a "simple" chatbot for managing certain tasks. But I seem to lack the understanding of how (if possible) to add new intents on the go during runtime.
The issue I face is let's say I have the following in my yaml file:
type: intent
name: questionAboutFood
slots:
- name: foodType
entity: foodType
utterances:
- what color is a [foodType]
- where can I buy a [foodType]
---
type: entity
name: foodType
automatically_extensible: yes
values:
- banana
- apple
- orange
from this file I can fit it to my snips-nlu engine.
But How do I during runtime append more foodTypes?
CODE:
from snips_nlu import SnipsNLUEngine
from snips_nlu.default_configs import CONFIG_EN
import io
import json
seed = 42
engine = SnipsNLUEngine(config=CONFIG_EN, random_state=seed)
with io.open("dataset.json") as f:
dataset = json.load(f)
engine.fit(dataset)
parsing = self.engine.parse("what color is the apple?")

Answer 1

I can see two workarounds:
Re-train
One way to do that consists in dynamically re-training the engine with the dataset updated with the additional entity values. This may be a reasonable solution if your dataset is not too big, as the training will be quite fast.
Improve the dataset
The other solution consists in providing many more formulations of your intent in the yaml file, so that the resulting engine will be able to extract unseen entity values.
In your case, you only provided two formulations in which the entity is the last token. Besides, the entity values you provided are only unigrams. For these reasons, it is likely that the NLU engine will not be able to capture entities when they span several tokens and/or when they are located in the middle of the sentence.
Here is a slightly better dataset which will already generalize much better:
type: intent
name: questionAboutFood
slots:
- name: foodType
entity: foodType
utterances:
- what color is a [foodType]
- what is the color of an [foodType] in general
- how much does an [foodType] cost
- can you describe a [foodType] for me please
- in which country is [foodType] generally found
- where can I buy a [foodType]
- how big are [foodType] in average
---
type: entity
name: foodType
automatically_extensible: yes
values:
- banana
- apple
- orange
- red meat
- black eyed pea
- brussels sprouts