# US 2013/0232162 A1 · Document checking apparatus, computer-readable recording medium, and document checking method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2013/0232162 A1
- Document ID: US20130232162A1
- Status: Published application
- Kind: A1
- Application number: 13/873,635
- Filed: 2013-04-30
- Published: 2013-09-05
- Earliest claimed: 2009-03-02
- Assignees: Fujitsu Limited
- Inventors: Hiroshi TSUDA
- Web page: https://patentgrep.com/patent/US20130232162A1

## CPC

- G06F 16/93: Document management systems
- G06F 21/554: {involving event detection and direct action}
- H04L 51/212: using filtering or selective blocking
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

A document checking apparatus includes a keyword appearance position extracting unit that extracts keywords and the appearance positions of the keywords from a target document including a confidential document; a keyword pair extracting unit that treats each keyword of the appearance positions of the extracted keywords as a target and determines whether there is another extracted keyword within a predetermined range before and after the target keyword; a feature element matrix creating unit that generates, when it is determined that there is the another keyword, combination information obtained by combining the determination target keyword and the another keyword in association with anteroposterior information of the appearance positions of the keywords; and a computing unit that determines whether the number of combination information, among the plurality of combination information of the generated target document, identical to the combination information of the confidential document is not less than a predetermined value.

## Claims

### Claim 1

A document checking apparatus comprising:
a memory; and
a processor coupled to the memory, wherein the processor executes a process comprising:
extracting character strings from a target document and a confidential document that becomes a comparison target when determining similarity between documents;
first determining, based on appearance positions of the character strings extracted at the extracting, whether another character string is within a predetermined range extending from a target character string toward either front or rear of the documents, to determine an element of feature element information corresponding to a pair of the target character string and the another character string within the predetermined range;
generating, when it is determined at the first determining that there is the another character string, the feature element information having a specific value in the element determined at the first determining;
comparing the feature element information of the target document and the feature element information of the confidential document generated at the generating; and
second determining whether a number of common elements having the specific value commonly among the feature element information of the target document and the confidential document is not less than a first similarity determination number to determine similarity between the target document and the confidential document on the basis of the determination result.

### Claim 2 (depends on 1)

The document checking apparatus according to claim 1, wherein the second determining includes determining, when it is determined that the number of the common elements is not less than the first similarity determination number, whether a number of a group of at least three or more character strings that appears within the same predetermined range is not less than a second similarity determination number.

### Claim 3 (depends on 2)

The document checking apparatus according to claim 2, wherein the process further comprises rejecting, when it is determined that the number of the group of at least three or more character strings that appears within the same predetermined range is not less than the second similarity determination number, transmission to another transmission partner other than a transmission permit partner of the target document.

### Claim 4 (depends on 1)

The document checking apparatus according to claim 1, wherein the memory stores the feature element information of the target document generated at the generating as the feature element information of the confidential document when the target document is transmitted to a transmission permit partner.

### Claim 5 (depends on 1)

The document checking apparatus according to claim 1, wherein
the second determining includes:
determining, when it is determined that the number of the common elements is not less than the first similarity determination number, whether an appearance density of the pair of the target character string and the another character string inside the target document is higher than a reference value; and
specifying, when it is determined that the appearance density of the pair of the target character string and the another character string inside the target document is higher than the reference value, a part that has a high appearance density by the determination as a similar part similar to the confidential document inside the target document.

### Claim 6 (depends on 5)

The document checking apparatus according to claim 5, wherein
the second determining includes treating a predetermined number of consecutive character strings included in the target document as a determination part and determines whether a number of pairs of the target character string and the another character string inside the determination part is not less than the reference value, and
the specifying includes specifying, when it is determined that the number of the pairs of the target character string and the another character string is not less than the reference value, the determination part as a similar part similar to the confidential document inside the target document.

### Claim 7 (depends on 6)

The document checking apparatus according to claim 6, wherein the second determining includes
treating, when the similar part is not specified, a predetermined number of consecutive character strings starting from a character string next to a forehand character string of the determination part as a next determination part,
treating, when the similar part is specified, a predetermined number of consecutive character strings ending at a character string corresponding to the common elements and located more backward than the specified similar part as a next determination part, and
determining whether the number of pairs of the target character string and the another character string inside the next determination part is not less than the reference value.

### Claim 8 (depends on 5)

The document checking apparatus according to claim 5, further comprising an output unit that outputs the similar part specified by the processor.

### Claim 9 (depends on 5)

The document checking apparatus according to claim 5, further comprising an optical character recognition unit that performs optical character recognition on a document image, extracts the recognized characters, and treats the extracted characters as the target document.

### Claim 10

A non-transitory computer readable storage medium having stored therein a document checking program, the document checking program causing a computer to execute a process comprising:
extracting character strings from a target document and a confidential document that becomes a comparison target when determining similarity between documents;
first determining, based on appearance positions of the character strings extracted, whether another character string is within a predetermined range extending from a target character string toward either front or rear of the documents, to determine an element of feature element information corresponding to a pair of the target character string and the another character string within the predetermined range;
generating, when it is determined that there is the another character string, the feature element information having a specific value in the element determined at the first determining;
comparing the feature element information of the target document and the feature element information of the confidential document; and
second determining whether a number of common elements having the specific value commonly among the feature element information of the target document and the confidential document is not less than a first similarity determination number to determine similarity between the target document and the confidential document on the basis of the determination result.

### Claim 11 (depends on 10)

The non-transitory computer readable storage medium according to claim 10, wherein
the second determining includes determining, when it is determined that the number of the common elements is not less than the first similarity determination number, whether a number of a group of at least three or more character strings that appears within the same predetermined range is not less than a second similarity determination number.

### Claim 12 (depends on 10)

The non-transitory computer readable storage medium according to claim 10, wherein
the second determining includes
determining, when it is determined that the number of the common elements is not less than the first similarity determination number, whether an appearance density of the pair of the target character string and the another character string inside the target document is higher than a reference value, and
specifying, when it is determined that the appearance density of the pair of the target character string and the another character string inside the target document is higher than the reference value, a part that has a high appearance density by the determination as a similar part similar to the confidential document inside the target document.

### Claim 13

A document checking method, comprising:
extracting character strings from a target document and a confidential document that becomes a comparison target when determining similarity between documents;
first determining, using a processor and based on appearance positions of the character strings extracted, whether another character string is within a predetermined range extending from a target character string toward either front or rear of the documents, to determine an element of a feature element information corresponding to a pair of the target character string and the another character string within the predetermined range;
generating, when it is determined that there is the another character string, the feature element information having a specific value in the element determined at the first determining, using the processor;
comparing, using the processor, the feature element information of the target document and the feature element information of the confidential document; and
second determining, using the processor, whether a number of common elements having the specific value commonly among the feature element information of the target document and the confidential document is not less than a first similarity determination number to determine similarity between the target document and the confidential document on the basis of the determination result.

### Claim 14 (depends on 13)

The document checking method according to claim 13, wherein
the second determining includes
determining, when it is determined that the number of the common elements is not less than the first similarity determination number, whether an appearance density of the pair of the target character string and the another character string inside the target document is higher than a reference value, and
specifying, when it is determined that the appearance density of the pair of the target character string and the another character string inside the target document is higher than the reference value, a part that has a high appearance density by the determination as a similar part similar to the confidential document inside the target document.
