how do you get bixby to recognize true/false concepts in training utterances?

I am making a capsule that gathers what the user says to run a real estate search. I am having trouble getting the training to recognize the true/false values though, like if the user wants a parking spot.

I tried highlighting the reference to parking, selecting the Boolean node, and writing "true" in the form section, but this does not seem to work.

Here is my training utterance:

"[g:SearchResultText] find (2)[v:Bedrooms] bedroom houses in (14850)[v:Zipcode] with (parking)[v:Parking:true]. my budget is (400000)[v:Price]"

However, it will not learn the training utterance. Any ideas what might be going wrong?

Solution

It is confusing for Bixby to take boolean value as input in this case. "find 2 bedroom houses in 14850 with no parking" would actually require parking=false.

I would suggest the following steps in this capsule.

Create an Enum premitive with predefined symbol such as "parking", "basement" to help filter your search. At this point, these keyword value can only be none or true. Does not support utterance such as "find me some property with no parking"
If decide to support such negative filter. You need another Enum to list all the conditions and train Bixby to understand "with no" or "without" speech. But I would think this step is for much later improvement or follow-up questions in the search.
To help Bixby understand "parking", "parkings", "parking space" are the same concept, you can use vocabulary rather than list all in your training. This link might be helpful. https://bixbydevelopers.com/dev/docs/dev-guide/developers/training.vocabulary#vocabulary-for-synonyms

Let us know if it helps, and have fun with Bixby!