US 11,868,421 B1
System and method for evaluating hyperdocuments using a trained artificial neural network
Ellis Robinson Giles, Houston, TX (US)
Filed by Ellis Robinson Giles, Houston, TX (US)
Filed on Jun. 19, 2018, as Appl. No. 16/012,745.
Application 16/012,745 is a continuation in part of application No. 29/604,162, filed on May 15, 2017, abandoned.
Application 29/604,162 is a continuation in part of application No. 14/537,892, filed on Nov. 10, 2014, abandoned.
Application 29/604,162 is a continuation in part of application No. 29/473,915, filed on Nov. 26, 2013, abandoned.
Application 29/473,915 is a continuation in part of application No. 12/984,594, filed on Jan. 4, 2011, granted, now 8,595,163, issued on Nov. 26, 2013.
Application 14/537,892 is a continuation of application No. 12/497,543, filed on Jul. 2, 2009, abandoned.
Application 12/984,594 is a continuation in part of application No. 12/103,446, filed on Apr. 15, 2008, granted, now 7,865,453, issued on Jan. 4, 2011.
Application 12/103,446 is a continuation of application No. 11/187,126, filed on Jul. 22, 2005, granted, now 7,403,929, issued on Jul. 22, 2008.
Claims priority of provisional application 61/133,740, filed on Jul. 2, 2008.
Claims priority of provisional application 60/590,812, filed on Jul. 23, 2004.
This patent is subject to a terminal disclaimer.
Int. Cl. G06F 7/00 (2006.01); G06F 16/955 (2019.01); G06N 3/02 (2006.01); G06F 16/9535 (2019.01)
CPC G06F 16/9558 (2019.01) [G06F 16/9535 (2019.01); G06N 3/02 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system for determining the disposition of a hyperdocument using a trained artificial neural network, the system comprising:
an information source component, wherein the information represents RGB pixel color data;
a content transfer component;
a component requesting information from the information source component through the content transfer component; and
a computer server component having one or more processors containing a trained artificial neural network having one or more hidden nodes and one or more hidden layers capable of evaluating the information from the information source component with the computer server configured to:
receive RGB pixel color data from the content transfer component;
dither the RGB pixel color data to a color palette;
create inputs to the artificial neural network using the dithered pixels;
evaluate the artificial neural network using a sigmoid activation function;
categorize the output value from the evaluation of the artificial neural network;
and return data to the content transfer component based on the category of the output value.