A python script for AI

[EdEarl]

I believe there are many things scientists have done and will do that are philosophical while trying to expand the limits of science, especially in the soft sciences such as psychology, neuroscience and linguistics.

[fredreload]

On 12/22/2017 at 2:09 AM, Strange said:

That would mean it wasn't science!

I am fairly sure that other models of language will turn out to be more successful and that, in a generation or so, Chomsky's work will be of only historical interest. Unfortunately, I won't be around to find out!

I can't help but think that getting a good dictionary is the core of this project. But since I didn't build this dictionary, I am referring to the unknown.

[Strange]

1 hour ago, fredreload said:

I can't help but think that getting a good dictionary is the core of this project. But since I didn't build this dictionary, I am referring to the unknown.

I think your basic idea of looking at the frequency of particular combinations words, is sound. That is how babies learn the sounds of language, where word-breaks are and the meaning and grammar of the language. So good luck with it.

[fredreload]

4 minutes ago, Strange said:

I think your basic idea of looking at the frequency of particular combinations words, is sound. That is how babies learn the sounds of language, where word-breaks are and the meaning and grammar of the language. So good luck with it.

Thanks Strange. But how often do you read from an article that says, "To eat is to put something into the mouth." If I have to compile a dictionary for verbs such as eat from is, am, are. And from a Wiikipedia dump. What source should I look into besides Wikipedia dump?

Edited December 24, 2017 by fredreload

[Strange]

5 minutes ago, fredreload said:

Thanks Strange. But how often do you read from an article that says, "To eat is to put something into the mouth." If I have to compile a dictionary for verbs such as eat from is, am, are. And from a Wiikipedia dump. What source should I look into besides Wikipedia dump?

Well, I suppose all sources. After all, children don't learn the meanings of words by reading encyclopaedias! They pick up the meaning of "eat" from the many different contexts they see it in. A story where someone sits down to eat dinner. Their grandma telling them to come and eat breakfast. And being told "eat this". They have the advantage of being in the physical world and so being able to associated objects and activities with the words, but it should work anyway. 

[fredreload]

59 minutes ago, Strange said:

Well, I suppose all sources. After all, children don't learn the meanings of words by reading encyclopaedias! They pick up the meaning of "eat" from the many different contexts they see it in. A story where someone sits down to eat dinner. Their grandma telling them to come and eat breakfast. And being told "eat this". They have the advantage of being in the physical world and so being able to associated objects and activities with the words, but it should work anyway. 

Right but I am associating eat based on its definition, but not by what is associating with it. For instance, "To eat" is "to put something into the mouth". If I have to remember it based on what can be eaten, it would be a huge list. For example, I can eat "a pineapple pie", "fruit", "cereal", etc. This would be a false lead

[Strange]

2 hours ago, fredreload said:

Right but I am associating eat based on its definition, but not by what is associating with it. For instance, "To eat" is "to put something into the mouth". If I have to remember it based on what can be eaten, it would be a huge list. For example, I can eat "a pineapple pie", "fruit", "cereal", etc. This would be a false lead

But that doesn't give you anything much more than a dictionary, does it? If you want your program to understand all the complex associations with the word "eat" (eating crow, eating your own dog food, you are what you eat, have your cake and eat it, etc.) then it needs to have more than just a simple definition.

[fredreload]

It seems you are going on a different approach to machine learning based on association. Not one I could comprehend at this point. But it is an open topic, so feel free to present one if you get it to work

[fredreload]

On 12/25/2017 at 1:08 AM, Strange said:

But that doesn't give you anything much more than a dictionary, does it? If you want your program to understand all the complex associations with the word "eat" (eating crow, eating your own dog food, you are what you eat, have your cake and eat it, etc.) then it needs to have more than just a simple definition.

My bad I'm being too harsh. How about we search the article by if statements? Like if this, do that. There's plenty of that in the Microsoft QA