Device for and method of measuring similarity between sets
AssigneeThe United States of America as represented by the Director, National Security Agency
InventorsSteven W. Knox, Richard M. Bates
Life of this patent
Estimating the term…
Abstract
The present invention is a device for and method of measuring similarity between sets using a union block, three function blocks, an adder, a subtractor, and a divider. The first set is fed into the first function block and the first input of the union block. The second set is fed into the second function block and the second input of the union block. The output of the union block is fed into the input of the third function block. The outputs of the first and second function block feed into the adder. The output of the adder and the output of the third function block feed into the subtractor. The output of the subtractor feeds into the input of the divider.
Claims25
A device for measuring similarity between a first set of items and a second set of items, comprising:
- a) a union block for performing a union function, having a first input for receiving the first set of items, a second input for receiving the second set of items, and having an output;
- b) a first function block for performing a user-definable non-linear function, having a first input connected to said first input of said union block, having a second input for receiving a dictionary of terms, and having an output;
- c) a second function block for performing the user-definable non-linear function, having a first input connected to said second input of said union block, having a second input connected to said second input of said first function block, and having an output;
- d) a third function block for performing the user-definable non-linear function, having a first input connected to said output of said union block, having a second input connected to said second input of said first function block, and having an output;
- e) an adder, having a first input connected to said output of said first function block, having a second input connected to said output of said second function block, and having an output;
- f) a subtractor, having a first input connected to said output of said third function block, having a second input connected to said output of said adder, and having an output;
- and
- g) a divider, having a first input connected to said output of said third function block, having a second input connected to said output of said subtractor, and having an output.
A method of measuring similarity between a first set of items and a second set of items that does not require any apriori knowledges of the first set of items or the second set of items, comprising the steps of:
- a) receiving the first set of items and the second set of items and performing a first user-definable function on the first set of items and the second set of items on a union block that performs a union function;
- b) receiving a dictionary of terms and performing a first user-definable, non-linear function on the result of step (a) and the dictionary of terms on a first function block;
- c) performing a second user-definable, non-linear function on the first set of items and the dictionary of terms on a second function block;
- d) performing a third user-definable, non-linear function on the second set of items and the dictionary of terms on a third function block;
- e) adding the results of step (c) and step (d) on an adder;
- f) subtracting the results of step (a) from the results of step (e) on a subtractor;
- and
- g) dividing the result of step (f) by the result of step (b) on a divider and returning the result as the measure of similarity between the first set of items and the second set of items.
Dependencies are inferred from the claim text.