Pattern extracting apparatus and method

A pattern extracting method of extracting, from a plurality of items included in target information pieces, a pattern of a combination of two or more of the items different from each other, the method comprising:

• the step of storing the plurality of target information pieces in a storage area;
• the step of producing candidate patterns each formed of two or more of the items different from each other based on each of the items included in each of the plurality of target information pieces;
• the step of storing an association degree between the items in a storage area,
• the step of calculating an extraction evaluation value of the candidate pattern based on a frequency of appearance at which the produced candidate pattern appears in each of the plurality of target information pieces;
• and
• the step of determining any of the candidate patterns having the calculated extraction evaluation value satisfying a predetermined threshold value and extracting the candidate pattern satisfying the threshold value,
• wherein the step of calculating the extraction evaluation value of the candidate pattern includes identifying the association degree between the items included in the candidate pattern and calculating the extraction evaluation value based on a weight based on the identified association degree and the frequency of appearance.

A non-transitory computer readable medium including computer executable instructions causing a computer to perform functions, the computer including a predetermined storage area storing target information pieces including a plurality of items and an association degree between the items, the computer performing pattern extraction processing of extracting a pattern of a combination of two or more of the items different from each other existing in the target information pieces, the functions comprising:

• the function of producing candidate patterns each formed of two or more of the items different from each other based on each of the items included in each of the plurality of target information pieces;
• the function of calculating an extraction evaluation value of the candidate pattern based on a frequency of appearance at which the produced candidate pattern appears in each of the plurality of target information pieces;
• and
• the function of determining any of the candidate patterns having the calculated extraction evaluation value satisfying a predetermined threshold value and extracting the candidate pattern satisfying the threshold value,
• wherein the function of calculating the extraction evaluation value of the candidate pattern includes identifying the association degree between the items included in the candidate pattern and calculating the extraction evaluation value based on a weight based on the identified association degree and the frequency of appearance.