Terms not really filtering anything

[alexaand]

After i answer something like "we need tomatos" at a question about an ingredient in a recipe, what i find in terms from the nlu is "we need tomatos". This however is not the proper ingredient. Shouldn't "we need" be cut off? Does the nlu always save the whole sentance on first position (maybe this is why I always get the full sentance?)

[narigaoka]

i haven't thought of the case "we need ___". i have added it. can you try again? and if you find other cases you can add them here. I sure haven't thought of all possible cases. Only those which seemt common to me.

[rosine]

yes this works now

[narigaoka]

really? i just noticed i forgot a net, so term.gra actually shouldn't have been able to compile ... i added the missing net. now it should be working

[rosine]

yes it did work, but another case didn't work which works now

[alexaand]

for keywords: (can you) help me (can you) assist me

[narigaoka]

you mean the keywords now and not term, right?

[rosine]

well, probably because those are no keywords yet^^

[rosine]

would those be for KA, RA and CI?

[narigaoka]

term is currently majorly filtering words for recipe learning. if you need terms for other cases, add them here, please.

[xixilian]

I can split the term in CI self...it's okay

[xixilian]

sorry maybe it's better that CI and CR can get the norms, like "what's the price of spaghetti napoli" and just "spaghetti napoli" in terms ...it would be nice for us

[narigaoka]

i will add these cases. but could it be that someone saved "spaghetti napol" (napoli without i) as keyword?

[narigaoka]

i have added the cases:

• can you tell me the price of _____
• what is the price of ____

The robot says: I've found your user history, Meng Meng. How may I help you Your next Input:can you tell me the price of pizza quattro formaggi pizza quattro formaggi // <- in term can you tell me the price of pizza quattro formaggi // <- in term Keyword 0: price dm.CanteenInformationState The robot says: The price for wanted meal it's 2.0 Your next Input:

the price (2.0) is wrong, though

[rosine]

all I found was "spags" in recipeData

[xixilian]

now it will show the right price...but output not always right

[alexaand]

"The fisrt/next ingredient/tool is (term)" is not filtetered by nlu

[Germandrummer92]

Would be very helpful if you could add the following:

bring me (term) where is (term) help me find (term)

Thanks!

[narigaoka]

Your next Input:can you bring me the car Keyword 0: bring Term 1: car Term 2: can you bring me the car The robot says: I need your name for getting your user history. You know my name after all! Your next Input:bring me ducks Keyword 0: bring Term 1: ducks Term 2: bring me ducks The robot says: Sorry. I don't know where ducks is, or I don't know how to do that Your next Input:where is the door Keyword 0: where is Term 1: door Term 2: where is the door Exception in thread "main" java.lang.NullPointerException at dm.CanteenInformationState.getOutputKeyword(CanteenInformationState.java:56) at nlg.OutputCreator.createOutput(OutputCreator.java:49) at nlg.NLGPhase.setPhaseResult(NLGPhase.java:21) at generalControl.Main.startDialog(Main.java:53) at generalControl.Main.main(Main.java:285) i have edited the grammar but an exception occured in the canteen dialog.