Watson Assistant Dialog Skill如何区分具有相同同义词的两个实体值？

Question

I'm looking forward a solution to the problem when several Entity Values have the same Synonyms. I will like to train Watson Assistant so that it goes back to the user and ask for the correct entity value under this ambiguity. How can we train Watson Assistant or what capability should we use to solve this ambiguity between entity Values?

Currently, I have an entity called @enterprise_name. We will have the following Entity Values with the same matching Synomyms:

Entity Value | Synonym |

EnterpriseNameExample 1 | EnterpriseNameExample |

EnterpriseNameExample 2 | EnterpriseNameExample |

EnterpriseNameExample 3 | EnterpriseNameExample |

• If the user asks: Can you give me information about EnterpriseNameExample?

• Currently, the watson assistant will either match EnterpriseNameExample to EnterpriseNameExample 1 or EnterpriseNameExample 2 or EnterpriseNameExample 3

• I will that the watson assistant gets the match with EnterpriseNameExample 1 and EnterpriseNameExample 2 and EnterpriseNameExample 3 and goes back to the user with the question: What Enterprise did you mean: EnterpriseNameExample 1, EnterpriseNameExample 2, or EnterpriseNameExample 3? and the user to select one.

As in our data they are considered as 3 different enterprises, but user could go and ask just for the short name.

Answer 1

If you want to disambiguate EnterPriseNameExample that can belong to one of the three categories I would suggest to use following pattern:

EnterpriseNameExample 1 | some non-ambiguous examples |

EnterpriseNameExample 2 | some non-ambiguous examples |

EnterpriseNameExample 3 | some non-ambiguous examples |

EnterpriseNameExample 1_2_3 | EnterpriseNameExample |

now when the entity detects EnterpriseNameExample 1_2_3 you know it is time to ask the users whether he/she meant 1, 2 or 3.

The system in general cannot handle multiple entity values with the same synonyms.