readme.txt

/**
 * @author jeffreymeyerson
 * 
 * 
 *         This experiment tests the results of a data set being put through a
 *         system modeling a dating website. The data set initializes a User set
 *         and defines their UndisclosedPreference set. The User set has been
 *         generated by a Python program which does the following: 1) Define a
 *         Trait set of single-value Traits. The same Trait will have different
 *         values across the population. 2) For each User, for each Trait
 *         defined in the previous step, declare that Trait to have the value of
 *         a random double between 0 and 10. 3) For each User, define an
 *         UndisclosedPreferences which is derived from Traits from the set
 *         created in 1). This Trait set is converted to a single trait through
 *         the TraitConvertible interface, and now belongs to the User's Trait
 *         set wihin a User's UndisclosedPreferences. 4) For each User, define
 *         that user's profile by assigning a random value to each Trait
 *         required by profile definition; this Trait set defines a User. 5) For
 *         each User, initialize that User's list of predicted sought traits
 *         with three things: a) a random Trait t from among that User's
 *         Preference set, b) the value of the Trait belonging to the numerator
 *         in t, and c) the value in the trait belonging to the denominator in
 *         t.
 * 
 *         The goal of the system is to provide relevant suggestions to each
 *         user correlative to that user's UndisclosedPreferences set while
 *         knowing as few of that user's UndisclosedPreferences explicitly as
 *         possible. In terms of a user's UndisclosedPreferences, the system
 *         creates an OptimalMatch vector for each user based on these
 *         Assumptions. It then uses vector similarity to figure out how closely
 *         each other user is to the OptimalMatch, before presenting a set of
 *         Users as a Suggestion.
 **/