7th April 2003
The Birth of Astroboy
I had heard from the start of the year that the
cartoon character called Astroboy who had been invented a few
decades ago was said to be born on this date. I remember watching
the series as a young child in the late '80s. Eventhough in
reality today we are only beginning to discover how to build
smart robots without touching on complete artificially
intelligent systems.
To offer my tribute to the little guy I prepared a program using
what I had already constructed from my natural language analyser
to work as a self-learning program which acquires the language
typed into it. It acquires this language the same way a baby does
(from scratch). This program called Astroboy was compiled at 12:02:44
AM today and not long after that he has been building up his
lexicon or vocabulary. This program not only learns new words but
also builds some basic semantic relationships with the words he
hears. These semantic relationships are not saving to a file yet.
Also around the corner I hope to build procedures that, with the
use of a microphone, will perform voice recognition and then the
Astroboy program can build it's own pronunciations.
Note how I refer to Astroboy as him, it is by personal convention
that I use this pronoun.
3rd March 2003
Announcement on the Development of my
Natural Language Analyser
For the past 2 years now I have researched and
developed the concepts of natural language processing drawing on
the research on how the human mind does this same task. Comprised
in these 2 years was an initial 1 and a half years research and
the last half year in development using the C language and soon
to follow in C++. As of today I have a completed computer program
capable of the following tasks for Natural Language Analysis:
* Segments a strings of characters into words and if possible
segments those words into prefixes, suffixes and base word
compounds. This especially works for additional consonants and
'e's.
* Stores the words into phrase structure which are stored in
sentence structures. This allows for manipulation for example
restructuring what was read from question into answer or vice
versa.
* Learns new words and saves them into it's personal dictionary.
Not only are new words remembered by it's spelling but also by
it's class (eg Noun, Verb, etc) and therefore where it should
appear in a sentence.
* Associates words to other words based on past learning. The
program will pick up on statements being true, false and
contradictive based upon classical conditioning described by
phycologists.
Useful applications to follow will be:
Spelling and Grammar Checker - self explanatory
Internet Censorship Program - Selectively blocks sites
depending on it's language content.
Knowledge Base Systems - Something of a database except
the computer takes the task of designing the structure while the
user only has to enter data and ask queries.
Expert Systems - Such a program works as backup or
substitute to someone of a particular profession. For example if
a computer technician is unavailable an expert system in computer
repairs would be helpful.
Universal Language Translator - Using a base language,
most likely English, this program should be able to translate one
language into another (word for word, phrase for phrase and
context for context).