|
behnamy - 2015-09-01 18:03:50
Hi Samuel, I read your article, it was great! and really thank you.
I have a question about artificial intelligence that I think you worked in this field before and you can help me.
I want to code a script that can analyze a sentecne/text as parameter like yours, but my goal is to define that the parameter is well written text in English/or other languages. I don't mean defining spelling of words or grammar mistakes, I mean to check if is it a meaningful sentence that the english native speakers say it or not. for example look at these two sentences:
1- I am going to the university tommorrow.
2- I am gasgas to sag sgasg bngasg.
as you can see the first sentence is right, but none of the English native speakers would say the second sentence!
or another example:
1- I am going to university tommorrow.
2- I am going to university coming.
as you can see the "coming" word is wrong after university!
so for achieving this goal(anyalizing text for checking the right sentece in multiple ways) I have some questions about it:
1- Is there any available library/code that I can use it?
2- I think this project is related to artificial intelligence, what is the best way/reference to study about it? (I want to use it in PHP)
3- how much hardware resources do I need to do this project well? someone told me that I need at least one petabyte of space for saving the sample texts for training the machine :|
Samuel Adeshina - 2015-09-01 18:26:23 - In reply to message 1 from behnamy
Hi Behnamy,
The solution to the problem you are trying to solve is a "complete subset" of artificial intelligence but it is completely different from the concept of AI.
It is called natural language processing and I've done something like that in java (if you are familiar with java, then please check here: http://github.com/Samshal). It is an huge field of study where you focus on things like Machine translation, Language Modeling, string Lemmatization, part of speech tagging, computational linguistics, text Parsing and so on.
If what you need is a quick solution for a problem you have at hand, you can easily check out this library: http://php-nlp-tools.com/ . It is a widely used library for natural language processing in php, it has a huge community behind it with a lot of support.
But if you are looking into doing a research, then you may be in luck cause I'm also working on something similar currently
check here: http://github.com/Samshal for a part-of-speech tagging and string lemmatization project I'm working on, it's still been developed but the documentation is kind of comprehensive enough
Or send an email to <samueladeshina73@gmail.com> so I can share some of my findings with you and we'ld talk more about algorithms for natural language processing
Samuel Adeshina - 2015-09-01 18:33:10 - In reply to message 1 from behnamy
P.S: i almost forgot :)
Machine learning, data processing, Artificial intelligence and so on are not scary fields!!!
"Someone" is trying to scare you away by saying "You'd need about a petabyte of memory"
... that's some big misconception people who doesn't have any idea about this concepts try to muddle up to scare away guys like you and I.
Just set your mind to it and you'd find out that with proper optimization and the right algorithms you wouldn't even need upto a gigabyte of harddisk to get acquainted with and opinionated about this cutting-edge fields
|