Switching to rive really paid off. The database for personality is growing at an alarming rate. The total size at the time of this writing is 14.6GB. I am having small issues with the data array input into the personality correctly, causing odd false positives. Example: “Milo, what is the internet?” Return: “I do not know what BEEF is.”… I will be diving into this issue soon.
Atfer heavy research with AIML (artificial intelligence markup language), and not really finding what I needed in order to make LIBBY/MILO a little more.. well..
life like.. I finally found a decent alternative to aiml. It is called rivescript by Casey Kirsle. (Guess they were also fed up with AIML and it’s odd implementation array’s).
for creating a perfect solution for the brain.
I am working on mixing the array structure & data mining of MILO into RiveScript to provide unique vocal response to ‘personal’ questions asked.
MILO’s personality will be changing based on the questions he is asked, and the data he pulls down from the questions..
#easch - eric@purpletree.org
#corebrain - fnabuild: ae333ed73jddl-cb
A structure example of LIBBY+MILO+RiveScript:
Human(VOCAL): Milo Search “LISP” <-----------
|
LIBBY(passes vocal data to MILO) :NEXT <----|----------------------------|[TAKE VOICE]
|------------ ^
| ^
MILO ->(check.local.db)/.times.spoken/asked :NEXT<------|--------- ^
| ^
MILO ->(check.remote.mit.START)/.compare-local/remote :NEXT <----| ^
MILO ->(download.return.data.from.questions) :NEXT <-------------| ^
| ^
MILO ->(PASS.VOCAL.DATA.BACK.TO.LIBBY) :NEXT <-------------------|-------^[PASS ANSWER]
| ^
MILO ->(build.more.search.items.from.last.response) :NEXT <------| ^
| ^
MILO ->(save.rogue.data.from.arbituary.search) :NEXT <-----------| ^
| ^
MILO ->(search.all.data.regex.to.rivescript.brain) :RETURN <-----|-------^[ADD TO PASS]
Metamorphasise Example of RS-Brain based on changes from the question “LISP?”
Question first ask:
// original // after MILO/LIBBY passthrough (q)+ what is lisp? Question second ask: (q)+ what is lisp? (GO TO MILO ->(check.local.db)/.times.spoken/asked :NEXT) Question third ask: (q)+ what is lisp? (GO TO MILO ->(check.local.db)/.times.spoken/asked :NEXT) As we can see from the above example, we take a basic search string, and speak it aloud. The string is then passed from LIBBY (listening) to MILO (mining) We take a string never asked before example: “LISP”; and MILO then searches for it in the wild. It pulls down what it finds. From the found data, it does searches The more times we ask about a specific subject, MILO will start to get more interested in the subject we are asking about; and by using rivescript, we
(q)+ what is lisp? (GO TO MILO ->(check.local.db)/.times.spoken/asked :NEXT)
(a)- i dont know
(a)-
--> insert <"LISP"> to search point system ‘1′ (one/zero -up to 20- point system)
(a)-
--> insert <"LISP"> to search point system ‘2′ (one/zero -up to 20- point system)
(a)- LiSP seem’s interesting to me! (second query, more interested) :SPEAK
(a)-
--> insert <"LISP"> to search point system ‘3′ (one/zero -up to 20- point system)
(a)- I like LISP. (third) query, more interested): SPEAK
(C) start entropy search on ‘lisp contents’ to build a stronger ‘LISP’ dataset.
and from MILO to RS.
based on the answer’s provided, and starts to build an internal mind from the data.
are now able to make it’s patterns a part of it’s personality.
