Document anonymity setting device, method and computer readable recording medium recording anonymity setting program

• US 20020091516A1
• Filed: 08/06/2001
• Published: 07/11/2002
• Est. Priority Date: 01/05/2001
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Ochitani, Ryo, Matsuoka, Hidesato
• Granted Patent: US 7,184,947 B2
• Time in Patent Office: Days
• US Class Current: 704/231
• CPC Class Codes: G06F 21/6245 Protecting personal data, e.g. for financial or medical purposes
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models

Abstract

A document anonymity setting device comprises a document input means inputting a document, a specificity calculating means extracting an expression specifying a person from the input document and for calculating a specificity to evaluate a degree of specificity at which the expression specifying a person, and an anonymity setting processing means rewriting with anonymity setting a expression in the input document having a specificity which is greater than a predetermined threshold. The specificity calculating unit extracts a person name and a modification expression from the input document and calculates a specificity to evaluate a degree of specificity at which the person name and modification expression thus extracted can specify a person. The anonymity setting processing unit rewrites a person name and a modification expression which have a specificity greater than a predetermined threshold through rewriting to meaningless expression, rewriting to low specificity setting, and rewriting to encrypted expression.

Claims

1. A device comprising:

a specificity calculating extracting means an expression specifying a person from the input document and for calculating a specificity to evaluate a degree of intensity at which the expression specifying a person; and

an anonymity setting processing means an expression specifying a person in the input document having a specificity which is greater than a predetermined threshold.

2. The device according to claim 1, wherein the specificity calculating unit extracts a person name from the input document and calculates a specificity to evaluate a degree of intensity at which the person name thus extracted can specify a person, and

the anonymity setting processing unit rewrites a person name having a specificity which is greater than a predetermined threshold.

3. The device according to claim 1, wherein the specificity calculating unit extracts a modification expression of a person name from the input document and calculates a specificity to evaluate a degree of intensity at which the modification expression thus extracted can specify a person, and

the anonymity setting processing unit rewrites a modification expression having a specificity which is greater than a predetermined threshold.

4. The device according to claim 1, wherein the specificity calculating unit includes:

a sentence extracting means extracting a sentence from an input document;

a part-of-speech analyzing means decomposing the extracted sentence for each part of speech;

a name extracting means extracting a person name based on a person name extracting rule from a result of the part-of-speech analysis; and

a name specificity calculating means calculating a specificity of the name thus extracted based on statistical information.

5. The device according to claim 1, wherein the specificity calculating unit further includes:

a syntax analyzing means creating a syntax tree indicative of a modification relationship between clauses based on a syntax analyzing rule from a result of a part-of-speech analysis;

a person specifying tree extracting means extracting a tree structure specifying a person as a person modification expression based on a person specifying tree extracting rule for the tree structure obtained by the parsing unit; and

a tree structure specificity calculating means calculating a specificity of the person specifying tree thus extracted based on statistical information.

6. The device according to claim 4 or 5, wherein the specificity calculating unit includes a reference specificity database registering specificity data having a set of an expression specifying a person, a type of a person name or a modification expression and a specificity which are created based on an existing document.

7. The device according to claim 4 or 5, wherein the specificity calculating unit includes a database creating means extracting a person name and a modification expression for each document from an existing document of a document database to calculate a specificity and for creating a reference specificity database registering specificity data having a set of an expression specifying a person, a type of a person name or a modification expression and a specificity.

8. The device according to claim 1, wherein the anonymity setting processing unit has a non-anonymity requiring database registering a non-anonymity requiring expression and a expression registered in the non-anonymity requiring database in an expression specifying a person extracted from an input document is not rewritten to anonymity expression.

9. The device according to claim 1, wherein the anonymity setting processing unit has an anonymity requiring database registering an expression to be rewritten to anonymity expression and all expressions in an input document registered in the anonymity requiring database are rewritten to the anonymity expression.

10. The device according to claim 1, wherein the anonymity setting processing unit rewrites an expression specifying a person extracted from an input document to no mean.

11. The device according to claim 1, wherein the anonymity setting processing unit replaces an expression specifying a person extracted from an input document to a low specificity expression having a specificity which is equal to or smaller than a threshold to be used for rewriting with an anonymity expression.

12. The device according to claim 1, wherein the anonymity setting processing unit encryptes an expression specifying a person extracted from an input document to carry out rewriting with an anonymity expression.

13. The device according to claim 15, further comprising a decrypting indicating means decrypting and displaying an encrypted expression when someone reads a rewrited document with an anonymity expression through encrypting by the anonymity setting processing unit.

15. The method according to claim 14, wherein the specificity calculating step extracts a person name from the input document and calculates a specificity to evaluate a degree of intensity at which the person name thus extracted can specify a person, and

the anonymity setting processing step anonymity sets a person name having a greater specificity than a predetermined threshold.

16. The method according to claim 14, wherein the specificity calculating step extracts a modification expression of a person name from the input document, calculates a specificity to evaluate a degree of intensity at which the modification expression thus extracted can specify a person, and the anonymity setting processing step sets a modification expression having a greater specificity than a predetermined threshold.

17. The method according to claim 14, wherein the specificity calculating step includes:

a sentence extracting step of extracting a sentence from an input document;

a part-of-speech analyzing step of discriminating the extracted sentence for each part of speech;

a person name extracting step of extracting a person name based on a name extracting rule from a result of the part-of-speech analysis; and

a person name specificity calculating step of calculating a specificity of a person name thus extracted based on statistical information.

18. The method according to claim 14, wherein the specificity calculating step further includes:

a syntax analyzing step of creating a syntax tree indicative of a modification relationship between clauses based on a syntax analyzing rule from the result of the part-of-speech analysis;

a tree structure specifying a person extracting step of extracting a tree structure specifying a person as a person modification expression based on a person specifying tree extracting rule for the syntax tree obtained at the syntax analyzing step; and

a tree structure specificity calculating step of calculating a specificity of the tree structure specifying a person thus extracted based on statistical information.

19. The method according to claim 17 or 18, wherein the specificity calculating step takes a weight average of a specificity obtained by a calculation of an expression which specify a person extracted from the input document with a specificity registered in a reference specificity database registering specificity data having a set of an expression specifying a person created based on an existing document, a type of a person name or a modification expression and a specificity, and normalizes the weight average.

20. The method according to claim 17 or 18, wherein the specificity calculating step includes a database creating step of extracting a person name and a modification expression for each document from an existing document of a document database to calculate a specificity and creating the reference specificity database registering specificity data having a set of an expression specifying a person, a type of a person name or a modification expression and a specificity.

21. The method according to claim 14, further comprising an anonymity setting indicating step of setting and changing a threshold to be used at the anonymity setting processing step.

22. The method according to claim 14, wherein the anonymity setting method retains, in a threshold database, a threshold used in an anonymity setting processing for each processing document and sets a last threshold as a default during the anonymity setting processing of a new input document.

23. The method according to claim 14, wherein the anonymity setting processing step does not replace a registered expression with reference to a non-anonymity setting requiring database in an anonymity setting expression extracted from an input document.

24. The method according to claim 14, wherein the anonymity setting processing step all expressions is rewritten with anonymity expression in an input document registered in an anonymity setting database registering an expression to be always subjected to rewrite with an anonymity expression.

25. The method according to claim 14, wherein the anonymity setting processing step replace an expression which specifies a person extracted from an input document into a rewriting to meaningless expression.

26. The method according to claim 14, wherein the anonymity setting processing step replaces an expression specifying a person extracted from an input document with a low specificity expression having a specificity equal to or lower than a threshold which is to be used for rewriting of the expressions specifying a person.

27. The method according to claim 14, wherein the anonymity setting processing step encodes an expression specifying a person extracted from an input document, thereby carrying out anonymity setting.

28. The method according to claim 27, further comprising a decrypting indicating step of decoding and displaying an encrypted expression specifying a person when reading an anonymity setting document which is subjected to the anonymity setting through the encoding at the anonymity setting processing step.

14. A method comprising:

a specificity calculating step of extracting an expression specifying a person from the input document and calculating a specificity to evaluate a degree of specificity at which the expression specifying a person; and

an anonymity setting processing of rewriting with an anonymity expression in the input document which has a greater specificity than a predetermined threshold.

29. A computer readable recording medium recording an anonymity setting program comprising:

a specificity calculating an expression specifying a person from the input document and calculating a specificity to evaluate a degree of intensity at which the expression specifying a person; and

an anonymity setting replacement a expression in the input document which has a greater specificity than a predetermined threshold.