An artificially intelligent communications apparatus, comprising:
- (a) a central processing unit;
- (b) a first database storing a numerically quantifiable first attribute, said first database being in communication with said central processing unit;
- (c) a second database storing a numerically quantifiable second attribute, said second database being in communication with said central processing unit;
- (d) a third database storing a numerically quantifiable third attribute, said third database being in communication with said central processing unit;
- (e) an input communications channel for receiving a plurality of bids in a particular auction for a particular property, each of said bids comprising a numerical value for the first, second and third numerically quantifiable attributes and bid identification information;
- (f) a non-volatile memory bearing a software program, said software program controlling said central processing unit:
- (i) to store, for each bid, said first, second and third numerically quantifiable attributes and their associated bid identification information together with information whether the bid was selected,
- (ii) for a particular auction select those bids in which said second and third numerically quantifiable attributes are the same or within a specified numerical range and store the same in a first normalized database,
- (iii) for a particular auction select those bids in which said first and third numerically quantifiable attributes are the same or within a specified numerical range and store the same in a second normalized database,
- (iv) for a particular auction select those bids in which said first and second numerically quantifiable attributes are the same or within a specified numerical range and store the same in a third normalized database,
- (v) for said first numerically quantifiable attribute determine a magnitude of difference in said first numerically quantifiable attribute causing a bid selection compared to less favorable rejected bids within said first normalized database by comparing bids within said first normalized database to each other to generate a first marginal driver;
- (vi) for said second numerically quantifiable attribute determine a magnitude of difference in said second numerically quantifiable attribute causing a bid selection compared to less favorable rejected bids within said second normalized database by comparing bids within said second normalized database to each other to generate a second marginal driver;
- and
- (vii) for said third numerically quantifiable attribute determine a magnitude of difference in said third numerically quantifiable attribute causing a bid selection compared to less favorable rejected bids within said third normalized database by comparing bids within said third normalized database to each other to generate a third marginal driver;
- and
- (g) comparing a selected bid from an agent to other bids on the same property to generate a suggested bid derived from a marginal driver and a numerically quantifiable attribute.