[Bragi]

User: I have 2 hands.
User: Each of my hands, has 5 fingers.
User: How many fingers do have?

This would almost work. I still have to do 'each', similar as 'both'. If you were to leave out that word and the comma, it would work. You basically declared the smallest amount of data that needs to be supplied to do the calculation.
I am currently at the stage that the most tricky part is, to get the sentences parsed correctly. This occasionally requires some manual adjustments.
To me, it becomes more interestingly when I can try out things like:
-a human has 2 hand, 5 fingers each -> multiple items not yet handled correctly
-i am a human
-i am missing 1 finger -> negation is not yet  handled correctly.
or: my left hands only has 4 fingers -> overwriting of abstract data in the concrete is not yet done
-how many fingers do i have.

[Bragi]

I must confess that although I have read your thread, it was difficult for me to understand what is actually going on.

Ye, the blog structure doesn't help since you see the end first. Perhaps it would be easier to start with the earlier posts. The application's help file also tries to explain some and the white paper explains yet another part.

[sunama]

To me, it becomes more interestingly when I can try out things like:
-a human has 2 hand, 5 fingers each -> multiple items not yet handled correctly
-i am a human
-i am missing 1 finger -> negation is not yet  handled correctly.
or: my left hands only has 4 fingers -> overwriting of abstract data in the concrete is not yet done
-how many fingers do i have.

The above would be interesting because at that point, your program is now "thinking" and assisting in finding solutions to problems.

I would love to know when you reach this stage.

[Bragi]

Thanks, yes it would be nice to get there. It shouldn't be to hard though. I have a clear view of what still needs to be done and how. It's just a matter of doing the work, and mostly avoiding the obstacles along the way. In the end it's still just playing with structure, that's all.

[Bragi]

@sunama: In case you're interested, AICI can now count stuff.  Thanks for the little statement sequence, it helped me debug some more.

Time now to think about the blog caption text, Mmmm

[sunama]

@sunama: In case you're interested, AICI can now count stuff.  

So specifically, what do you mean count stuff?

Can you give some examples of input/output? Perhaps you can update this information in your blog?

Also how does your AI respond to the following input:
User: I have green eyes.
User. What what colour are my eyes?

[Bragi]

already done, check http://janbogaerts.name/

[sunama]

already done, check http://janbogaerts.name/

Impressive. Your AI is now able to apply logic and not just regurgitate information.

Dealing with plurals really is a minor problem...its not a big deal.

Can you put through the following input and let me know what answer it gives:

User: I have green eyes.
User. What what colour are my eyes?

[Bragi]

I think the previous version can already do that. You can try the AICI client. In any case, I'll be releasing a new version of the client tomorrow or so, than you can try for yourself.

[sunama]

I think the previous version can already do that. You can try the AICI client. In any case, I'll be releasing a new version of the client tomorrow or so, than you can try for yourself.

I'm interested to know how you were able to associate the word "colour" with "green"?

My own AI is able to do this. It does this by using a word association value*.

The word association value is incremented every time a sentence is input, which has the words "colour" and "green" in it. Eventually, when enough sentences have been input containing the 2 words, the word association value reaches a high value, for the word pair (colour and green). From this point onwards, when AI receives a question, involving these 2 words and it can't answer the question using by running a basic search on previous user input, it resorts to looking for word associations.

I'm still on the look-out for another algorithm which allows the program to associate the words green and colour (obviously the word pair can be for any pair of words in the English dictionary; I am using "green" and "colour" as an example).

If you don't wish to the share your algorithm, then I will understand and won't be offended, but it would be nice to know how you have done it.

[Bragi]

First off, the algorithm that I use runs on my neural network application. This is more akin to a new breed of hardware processor (in fact, it's a virtualization of one). So it's pretty hard to do a similar algorithm in something like VB.net or C#.
That said, the association is done similar to how a thesaurus works. In fact, in the designer, this data is displayed (and can be edited) in the form of a thesaurus. When an unknown word is found that is parsed as the 'value' of something and the 'what' that got extracted out of the sentence is 'color', the network asks the word 'color' to handle this unknown (in other words, the word 'color' needs to tag and bag the word 'green'). 'Color' creates the thesaurus relationship and indicates that the word 'green' can be used as adjective and noun.  All of this only works because the network is able to parse a subset of the English language.

[sunama]

That said, the association is done similar to how a thesaurus works. In fact, in the designer, this data is displayed (and can be edited) in the form of a thesaurus.

Is this association done automatically or does the user have to feed the ai with the relationship between 'color' and 'green'.

When an unknown word is found that is parsed as the 'value' of something and the 'what' that got extracted out of the sentence is 'color', the network asks the word 'color' to handle this unknown (in other words, the word 'color' needs to tag and bag the word 'green'). 'Color' creates the thesaurus relationship and indicates that the word 'green' can be used as adjective and noun. 

How does aici know that 'colour' has to handle 'green'. Why doesnt aici choose 'animal' to handle the word 'green'? Does the user have to give extra input to hint that 'color' must bag and tag 'green'?

[Bragi]

Is this association done automatically or does the user have to feed the ai with the relationship between 'color' and 'green'.

This depends. Currently, when you say: my eyes are green, and 'green' is not yet known, it will always ask what green means. Even though you had previously said something like: 'we are know going to talk about colors'. Though this is technically fairly simple to add, I just haven't reached that far yet.
If you would have said: 'the color of my eyes is green', AICI would have been able to extract the relationship.

How does aici know that 'colour' has to handle 'green'.

Perhaps first a little bit about the network, you can look at it as a free form database. If you want to create a structure (record, class,...), you can have a neuron representing the object. Fields + values are done with links. The field name is represented by the link-meaning, the value is where the link points to. A table (list, collection,...) is done with clusters, which are simply specialized neurons, that can also be used as the object, value or field name.
With this in mind, Aici is a collection of neurons that mixes data and code. So, any word can have links to code blocks (also just neurons) (or can be used to execute something). When the Aici framework finds a 'what' for an unknown 'value' in a sentence, it checks if there is a code block attached to the word using a specific, known link-meaning. If this is the case, the code is executed. If no code block is found, it reverts to default behaviour: create an 'is a' thesaurus relationship, and tag the word with a part of speech. If it was a sentence like 'x is y', an adjective is presumed. If it is a sentence like 'x has y', a noun is presumed.

[sunama]

I understand now.

Thanks.

[sunama]

I've now worked your question and answers into my own algorithm (I havent coded it in yet, but after its done, my AI will also be able to answer the question you originally posted).

So now that your algorithm can multiply, can it divide?

What output do you get if you type in:
User: I have 10 fingers
User: A hand has 5 fingers
User: I have hands
User: how many hands do i have?

I haven't yet come up with an algorithm for the above, but I might do that tonight.