Curiosity in chatbot

[Sana Amanat]

1.Is curiosity an outcome of Emotion/Mood in some particular context? 

2. Is curiosity an emotion/Mood or something else?

In which circumstance a chatbot can show curiosity, what are the conditions? 

[Alexis Dimitriadis]

Sana,

I am puzzled by your questions. It's your chatbot and your model, so the
answers are up to you.

To state the obvious: Curiosity is an outcome if your model for
emotions/moods model treats "curiosity" as an emotion, i.e., if your
classifier or state machine generates it as a possible value. A chatbot
can show curiosity if it's programmed to behave that way (presumably by
asking questions). It's all up to you.

> 2. Is curiosity an emotion/Mood or something else?

That's an interesting question, if you ask a psychologist or a
philosopher. But for you, an emotion/mood is a state that determines how
your bot will behave. If you want it to act curious, well there you are.

Alexis

[Tim McNamara]

Hi Sana,

This question is probably too broad to be meaningfully answered.

Curiosity in a chatbot is a function of how the chatbot is
implemented. Attributing emotions and moods to a computer programme is
quite anthropomorphic.

There are countless ways to be define and implement something that
resembles curiosity.

Some things you will need to do:
- represent the state of the bot's knowledge somehow
- carry out analysis (introspection, of sorts) on that knowledge, to
identify gaps
- find some way of asking questions that will lead to answers that
are likely to fill the gap
- place a value judgement on the response (does it satisfy the need?)
- perhaps rephrase the question, or dig deeper

> --
> You received this message because you are subscribed to the Google Groups
> "nltk-users" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/nltk-users/-/37KA7did7KoJ.
> To post to this group, send email to nltk-...@googlegroups.com.
> To unsubscribe from this group, send email to
> nltk-users+...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/nltk-users?hl=en.

[Alexis Dimitriadis]

Should have started my answer by saying this: Treating curiosity this
way sounds reasonable; whether your bot does so, and how, is up to you.

Best,

Alexis

[Sana Amanat]

Well, Alex !! 

Actually my suggested model has a separate module of curiosity, which seems to be an overhead in some context, i want to integrate curiosity with emotion/mood module and want to treat curiosity like other emotions i.e sadness,happy etc,i want some solid arguments, i am bit confuse.Human mostly shows curiosity when s/he talks to some unknown person in some particular context, or when s/he want to know more about that.Can we map the same features in bots?   

[Sana Amanat]

Tim !! thanks

[richard kappler]

Sana, look at Suzette_the_Most_Human_Computer by Bruce Wilcox, the guy who created  ChatScript among other things.  While it's not nltk, if you read the entire article there's some very insightful ideas about doing what you seem to want to do. 

regards, Richard

On Sun, Aug 12, 2012 at 12:34 PM, Sana Amanat <sana.am...@gmail.com> wrote:

Well, Alex !! 

Actually my suggested model has a separate module of curiosity, which seems to be an overhead in some context, i want to integrate curiosity with emotion/mood module and want to treat curiosity like other emotions i.e sadness,happy etc,i want some solid arguments, i am bit confuse.Human mostly shows curiosity when s/he talks to some unknown person in some particular context, or when s/he want to know more about that.Can we map the same features in bots?   
 

--

"Treat all disasters as if they were trivialities but never treat a triviality as if it were a disaster." 
       -- Quentin Crisp