Step 1: Use a program to convert speech to text;
Step 2: Use another program to seperate the parts of speech (nouns and verbs are needed as far as i've planned this out);
Step 3: Using a third program + thesaurus database bring up a certain number of synonyms to the parts of speech detected;
Step 4: Match the noun(s) and the synonyms of the noun(s) to the listed commands;
a)If match found, continue to step 5;
b)If mulitple matches found, continue to step 5d;
c)If no matches found, have it access a database (lets say cleverbot) and talk back to you;
Step 5:Have the verb(s) and the synonyms of the verb(s) matched to the listed commands;
a)If matched, execute command;
b)If multiple matches, have it check the state of the program being imposed. If for example the music is already playing, have the command to play music disappear;
c)If multiple matches exist, have the system ask to specify and code the choice taken as the primary choice;
d)If mulitple nouns exist, have it do the same as 5b and try to throw out some options, if problem still persists have it do the same as 5c;
Those are the basics, some features I would like to implement are surroundings awareness, respond with text to speech with some commands, have the text to speech sound more human (there are programs for that).
Surroundings awareness: Have it know when i'm speaking to it or to someone else by methods of checking what programs are active(skype, some video playing etc...) and by knowing how many people are in the room with a face detection camera system.
I like tinkering with stuff and building things, but I never wanted to program or build a system untill now... I know the basics of programming and how it works (atleast I think I do) and I have a strong mathematical base. I would appreciate feedback on the idea, how complicated it is, methods of doing it, resources I can take advantage of to learn how to do it, any feedback really
Help me make Aries happen
