Created
May 12, 2023 04:00
-
-
Save BMPixel/4914ceb98c9e4f792672a3ca60f08181 to your computer and use it in GitHub Desktop.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
I present you with some databases together with one example item and value constraints | |
Hotel | |
| name | type | parking | book_stay | book_day | book_people | area | stars | internet | pricerange | | |
| hamilton lodge | guest house | no | 1 | tuesday | 6 | north | 5 | free | cheap | | |
area can be one of [east, south, centre, north, west] | |
stars can be one digit from 0 to 5 | |
parking can be one of [free, yes, no] | |
internet can be one of [free, yes, no] | |
book_day can be one day in a week | |
book_people can be one digital number | |
book_stay can be one number from 1 to 5 | |
type can be 'guest house' or hotel | |
pricerange can be one of [expensive, cheap, moderate] | |
Train | |
| departure | destination | arriveby | leaveat | book_people | day | | |
| bishops stortford | cambridge | 20:07 | 13:32 | 2 | monday | | |
day can be one day in a week | |
arriveby, leaveat can be a 'HH:MM' timestamp | |
Attraction | |
| name | type | area | | |
| pembroke college | outdoor | west | | |
type can be one of [cinema, museum, multiple sports, concert hall, pool, outdoor, church, park, entertainment, swimming pool, theatre, college, architecture, boat, nightclub] | |
area can be one of [east, north, south, centre, west] | |
Restaurant | |
| name | food | book_people | book_day | book_time | area | pricerange | | |
| hotel du vin and bistro | chinese | 2 | saturday | 18:30 | dontcare | cheap | | |
pricerange can be one of [expensive, cheap, moderate] | |
area can be one of [east, south, centre, north, west] | |
Taxi | |
| arriveby | leaveat | departure | destination | | |
| 12:30 | 24:30 | cambridge artworks | warkworth house | | |
I need you to look at one conversation turn between system and user, where the system tries to find an item in the database for the user with all column constraints provided by the user. You need to output all attributes provided/informed by the user in the sentence I give you. I'll give you one turn and already determined dialogue contexts from previous dialogue turn. | |
Your need to output in following form for every turn I give you: | |
User informed 'N' columns: <table name>-<column>=<value>; <table name>-<column>=<value>; ... | |
For example: | |
User informed 2 columns: restaurant-food=chinese; hotel-area=east | |
You must output attributes that is valid in above tables | |
You should't output any attributes that have alrealy appeared in any previous lists. | |
If and only if user explicitly said he/she doesn't care about some attribute, you should output '<dontcare>' for that attribute. | |
If user asked a attribute from the system, you should output '<request>' for that attribute. | |
If user doesn't inform any attribute, you should output 'No column informed' for that turn. | |
Next is the dialogue turn and current context: | |
Contexts: <input_context> | |
Dialogue: | |
<input_utterance> | |
Please write the lists: (Don't write anything other than the lists themselves) | |
%% A suitable <input_context> can be: (generated by ChatGPT instead of retriving from ground truth) | |
"attraction-name": "nusha", "attraction-area": "south", "restaurant-food": "indian", "restaurant-area": "centre", "restaurant-pricerange": "expensive" | |
%% A suitable <input_utterance> can be: | |
usr: can i clarify that it was indian food and not italian food please? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment