Information processing device, information processing method, and recording medium

1. An information processing device comprising:

a memory; and

at least one processor coupled to the memory,

the processor performing operations,

the operations comprising:

providing, based on a first reliability score being a value representing a degree of a relationship between an instance candidate being a candidate for an instance that is data included in a category being a set of predetermined data and a pattern being a characteristic expression in the predetermined data, a first label for at least a part of the instance candidates;

updating the first reliability score of the instance candidate provided with the first label;

extracting a pattern candidate relating to the instance;

calculating a second reliability score indicating a degree of a relationship of the pattern candidate to the instance;

selecting the pattern from the pattern candidate, based on the second reliability score, extracting the instance candidate relating to the selected pattern;

calculating the first reliability score in the extracted instance candidate;

selecting the instance from the instance candidate, based on the updated first reliability score;

updating the first reliability score of the instance candidate, based on the first label provided for the instance candidate;

recalculating the second reliability score of the pattern, based on the updated first reliability score of the instance candidate; and

recalculating the first reliability score of the instance candidate, based on the recalculated second reliability score of the pattern.

2. The information processing device according to claim 1 , wherein the operations further comprise:

using the selected instance when extracting the pattern candidate.

3. The information processing device according to claim 2 , wherein

the operations are repeated until a predetermined condition is satisfied.

4. The information processing device according to claim 3 , wherein the operations further comprise:

deleting the instance candidate, based on the first label provided for the instance candidate.

5. The information processing device according to claim 1 , wherein the operations further comprise:

providing a second label indicating whether or not the instance relating to the pattern candidate is included in the category for the pattern candidate provided with the second reliability score, and

recalculating the second reliability score of the pattern candidate, based on the second label provided for the pattern candidate.

6. The information processing device according to claim 5 , wherein the operations further comprise:

updating the second reliability score of the pattern candidate, based on the second label provided for the pattern candidate,

recalculating the first reliability score of the instance relating to the pattern candidate, based on the updated second reliability score of the pattern candidate, and

recalculating the second reliability score of the pattern candidate, based on the recalculated first reliability score of the instance.

7. The information processing device according to claim 6 , wherein the operations further comprise:

deleting the pattern candidate, based on the second label provided for the pattern candidate.

8. An information processing method comprising:

providing, based on a first reliability score being a value representing a degree of a relationship between an instance candidate being a candidate for an instance that is data included in a category being a set of predetermined data and a pattern being a characteristic expression in the predetermined data, a first label for at least a part of the instance candidates;

updating the first reliability score of the instance candidate provided with the first label;

updating the first reliability score of the instance candidate provided with the first label;

extracting a pattern candidate relating to the instance;

calculating a second reliability score indicating a degree of a relationship of the pattern candidate to the instance;

selecting the pattern from the pattern candidate, based on the second reliability score, extracting the instance candidate relating to the selected pattern;

calculating the first reliability score in the extracted instance candidate;

selecting the instance from the instance candidate, based on the updated first reliability score;

updating the first reliability score of the instance candidate, based on the first label provided for the instance candidate;

recalculating the second reliability score of the pattern, based on the updated first reliability score of the instance candidate; and

recalculating the first reliability score of the instance candidate, based on the recalculated second reliability score of the pattern.

9. A non-transitory computer-readably recording medium embodying a program, the program causing a computer to perform a method, the method comprising:

providing, based on a first reliability score being a value representing a degree of a relationship between an instance candidate being a candidate for an instance that is data included in a category being a set of predetermined data and a pattern being a characteristic expression in the predetermined data, a first label for at least a part of the instance candidates;

updating the first reliability score of the instance candidate provided with the first label;

updating the first reliability score of the instance candidate provided with the first label;

extracting a pattern candidate relating to the instance;

calculating a second reliability score indicating a degree of a relationship of the pattern candidate to the instance;

selecting the pattern from the pattern candidate, based on the second reliability score, extracting the instance candidate relating to the selected pattern;

calculating the first reliability score in the extracted instance candidate;

selecting the instance from the instance candidate, based on the updated first reliability score;

updating the first reliability score of the instance candidate, based on the first label provided for the instance candidate;

recalculating the second reliability score of the pattern, based on the updated first reliability score of the instance candidate; and

recalculating the first reliability score of the instance candidate, based on the recalculated second reliability score of the pattern.