Information processing device, method for verifying anonymity and medium
The present invention provides an information processing device that enables a reduction in the processing cost of verifying anonymity during anonymization when multi-dimensional data is the subject of anonymization. The information processing device is provided with: a unit which generates information indicating the correspondence between a record contained in a data set and a class specifying a unique combination of quasi-identifier attribute values; a unit which verifies the anonymity of each record on the basis of the class thereof indicated in the information; and a unit which, on the basis of the results of verifying the anonymity, updates the information in a manner such that whether or not the record satisfies the anonymity can be identified and outputs the record-class correspondence information.
1. An information processing device, comprising:
a CPU; and
a storage device including a program, wherein
the CPU reads the program from the storage device and executes function as:
a correspondence information generation unit which generates record-class correspondence information that indicates correspondences between a record included in a data set and a class that specifies unique combinations of quasi-identifier attribute values, which are values of quasi-identifiers included in said record;
an anonymity verification unit which verifies anonymity for each said record on the basis of said classes that are indicated in said record-class correspondence information; and
a correspondence information update unit which, on the basis of a result of verifying said anonymity, updates said record-class correspondence information in such a way whether or not that said record satisfies said anonymity is distinguishable and outputs said record-class correspondence information.
2. The information processing device according to claim 1 , the CPU reads the program from the storage device and further executes function as:
a current state generation unit which, on the basis of said record-class correspondence information and said quasi-identifier attribute values, generates, with respect to each variety of said quasi-identifier, current state information that indicates correspondences between said record, said class, and a recoded value corresponding to said quasi-identifier attribute value; and
a recoding unit which processes said recoded value included in said current state information,
wherein said anonymity verification unit, on the basis of said recoded value, verifies anonymity for each said record,
a current state update unit which, on the basis of said recoded value and a result of verifying said anonymity, updates said current state information and updates information corresponding to a result of the updating,
wherein said correspondence information update unit, on the basis of said updated current state information, updates said record-class correspondence information.
3. The information processing device according to claim 2 ,
wherein said recoding unit, on the basis of a generalization hierarchy that includes, at the highest level, a root with a value covering all values that said quasi-identifier can take and, towards lower levels, nodes with values that are specialized step-by-step, generates said recoded value.
4. The information processing device according to claim 1 ,
wherein said correspondence information update unit updates said record-class correspondence information by removing information corresponding to said record that does not satisfy said anonymity from said record-class correspondence information.
5. The information processing device according to claim 1 ,
wherein said correspondence information update unit updates said record-class correspondence information by appending information indicating that said anonymity is not satisfied to information corresponding to said record that is included in said record-class correspondence information and does not satisfy said anonymity.
6. The information processing device according to claim 1 ,
wherein said correspondence information update unit updates said record-class correspondence information by changing information corresponding to said record that is included in said record-class correspondence information and does not satisfy said anonymity to information indicating that said anonymity is not satisfied.
7. An anonymity verification method by which an information processing device executes the steps of:
generating record-class correspondence information that indicates correspondences between a record included in a data set and a class that specifies unique combinations of quasi-identifier attribute values, which are values of quasi-identifiers included in said record;
verifying anonymity for each said record on the basis of said classes indicated in said record-class correspondence information; and
updating said record-class correspondence information on the basis of a result of verifying said anonymity in such a way that whether or not said record satisfies said anonymity is distinguishable and outputting said record-class correspondence information.
8. The anonymity verification method according to claim 7 in which said information processing device executes the steps of:
generating, with respect to each variety of said quasi-identifier, current state information that indicates correspondences between said record, said class, and a recoded value corresponding to said quasi-identifier attribute value on the basis of said record-class correspondence information and said quasi-identifier attribute values;
processing said recoded value included in said current state information;
the verifying of said anonymity verifying anonymity for each said record on the basis of said recoded value;
updating said current state information on the basis of said recoded value and a result of verifying said anonymity and outputting information corresponding to a result of the updating; and
the updating of said record-class correspondence information updating said record-class correspondence information on the basis of said updated current state information.
9. A computer-readable non-transitory recording medium recording a program, the program making a computer execute the steps of:
generating record-class correspondence information that indicates correspondences between a record included in a data set and a class that specifies unique combinations of quasi-identifier attribute values, which are values of quasi-identifiers included in said record;
verifying anonymity for each said record on the basis of said classes indicated in said record-class correspondence information; and
updating said record-class correspondence information on the basis of a result of verifying said anonymity in such a way that whether or not said record satisfies said anonymity is distinguishable and outputting said record-class correspondence information.
10. The computer-readable non-transitory recording medium recording a program according to claim 9 , the program making a computer execute the steps of:
generating, with respect to each variety of said quasi-identifier, current state information that indicates correspondences between said record, said class, and a recoded value corresponding to said quasi-identifier attribute value on the basis of said record-class correspondence information and said quasi-identifier attribute values;
processing said recoded value included in said current state information;
the verifying of said anonymity verifying anonymity for each said record on the basis of said recoded value;
updating said current state information on the basis of said recoded value and a result of verifying said anonymity and outputting information corresponding to a result of the updating; and
the updating of said record-class correspondence information updating said record-class correspondence information on the basis of said updated current state information.