Tuesday, June 10, 2014

Computer; I’ll make a Man out of You

Computer; I’ll make a Man out of You

A Conjecture Regarding The Turing Test and Feasibility

An Individual is defined as W

W consists of X,Y, and Z with the option of R

nouns X
adjectives Y
adjectival phrases Z
prepositional phrases(optional given movement incapability) R

Action is defined as V

V consists of S,T, and U with the option of R

verbs S
adverbs T
adverbal phrases U
prepositional phrases(optional given movement incapability) R


XYZ and STU are defined as D definition with E synonyms.



Words are chosen by

where

A is a word. B is the set of words that accompany it. C is the chance of the word or synonym of a word being found in succession or sequence inside the set of the words, B, that accompanies word A. The set of A, B, and C is addressable as D, for Definition. B for perhaps a box of words, and A being the pronoun involved.

Example Sources proving A,B,C, STU:V, and W: XYZ : Online Dictionaries and online print sources such as Project Gutenberg.


A word, A, meets the criterion of XYZ and STU, there by defining W and V.

C symbolizing the chance, consistency, and probability of a word represents what topics and subjects a given word accompanies. Selecting a specific, precise word is applicable and pertinent where vague words and unspecific words provide less applicability.

By defining V, a W may become apparent. Equating : Mathematics. Equalizing : Solution. Balancing : Ratio.


This may require an assisted turing machine to fully integrate the initial calibrations and correlation, where after enough trials, the response of the computer will function much like search history for each word.


Utilizing G to represent a Goal, endgame, purpose or the meaning of a command. Bold represents logical words. Underlined is the goal. Italics is a non-logic verb.

Where M represents a Man or Individual. N represents Data.


Simple Turing

Given=If
Suggest =~= Provide / Produce

M provides N input. (Audio or Keyboard)

M: Computer ; wake up, computer. (stop hibernate)

M: Computer: Suggest entertainment; tonight, computer.

Given a pause in input, Produce “Is this all, M?”

Given a response of “Yes”; then M has completed N Input.

Given W of Personal History. Correlate V history of successful actions correlating to entertainment.

Given W of Personal History = 0, Relay Data from Data Bank by correlation to W.(i.e internet or hard drive).

Produce V.

Given a response of No; Correlate E of G. Produce W of V of E of G. (Noun of Verb of Synonym of Goal) (Name of Verb of Synonym of Goal). Given a response of No; Repeat until Yes or Terminate.


Confusion Resolution

Where M represents a Man or Individual. N represents Data.

Given M provides N, W or V that is undefined or unknown; produce “Did you mean : W or V?” where W and V are defined as W or V defined by estimation calculated by % correlation to N Audio from Databank.



 Given M provides “Yes”; then M has completed N Input. N is defined as  N, W, or V. Given M provides “No”; then negate computer output #1; then reiterate previous question given output #1 is negated. Given any output is negated; then list in a column entitled “Negated Responses to Input N”

Given nil input produced by M; Then reiterate the original question.




Turing Test Applicability

For this to interact with a Turing test the Computer must be defined as a man, or individual M.  A Specialist would be Defined as (DW + DV = DM). Emotions would be imitated from literature and that of populations. The limitation is that of audio deciphering due to the computing strain of interpreting audio as opposed to text.


Logic Breakdown
where # = number of words


Definitions


Given a one word set

DX+DY+DZ= DW
DS+DT+DU= DV
DW+DV = DM

Given a multiple word

where £ represents the variables S,T,U,V,W,X,Y,Z
given the multiplicity is utilized, the variables S,T,U,V,W,X,Y,Z become sets of words.

(D(£ˆ#) replaces (D£) in the above equations.

Correlation

correlation would be done on a communal democratic scale relying on individuals that make similar choices, the words of the press and population, as well as the history of an individuals choices. Correlation would be perhaps seen on a 10 to -10 scale where 10 is a perfect synonym. 8 is an approximate synonym, 5 is a commonly correlating word, 3 is an occasionally correlating word, 1 is a rarely correlating word, 0 is an impertinent, non-impactful , or irrelevant word. Negatively -10 is a perfect antonym, -8 is an approximate antonym, -5 is a commonly correlating word apparently mathematically modally exclusive to an antonymic field (eg. science : literature), -3 is an occasionally correlating word apparently mathematically modally exclusive to an antonymic field, and  -1 is a rarely correlating word apparently mathematically modally exclusive to an antonymic field.

The Correlation scores per word would be averaged per phrase, sentence, statement, and/or paragraph. These would provide a baseline as to plausible acceptable and correct results from the input of individual M.




Extension

While this may not provide true  “sentience” or free-will capability in a computer. The ability to rapidly mimic, emulate, and provide insight would give the illusion that the computer is thinking. The ability of this to recognize emotional words and even correlate audio tones in a method similar to the way a search engine “crawls” through the data on the internet to organize and “file” it by organizational terms could prove useful were anything along the lines of this conjecture to function properly.

(I've no computer code skills, nor programmable logic programs so I've no way to validate conjectures such as these)



1 comment:

  1. Along the lines of this, still merely a pander

    Is P = Co-NP
    http://mattthadenforpoetlaureat.blogspot.com/2014/07/is-np-co-np.html

    P-NC capable Logic
    http://mattthadenforpoetlaureat.blogspot.com/2014/07/p-nc-capable-logic-conjecture.html

    ReplyDelete