Confidential information identifying method, information processing apparatus, and program

摘要

An information processing apparatus includes a clustering unit configured to read messages from a log and to classify the read messages into clusters according to similarities of the messages; a variable portion finding unit configured to find a portion variable between messages; an attribute determination unit configured to estimate and determine a confidential attribute of the variable portion by using predefined rule; and an attribute estimation unit configured to, in a case where there is a portion whose confidential attribute is undeterminable by using the rules, estimate the confidential attribute of the portion having the undeterminable confidential attribute with use of either a correspondence between appearance locations in the messages, or a co-appearance relation of a portion having a determined confidential attribute and the portion having the undeterminable confidential attribute.

主权项

1. A method of allowing an information processing apparatus to identify confidential information in an accumulated log, the method comprising:
reading messages about operations of information processing apparatuses from the log, and classifying the read messages into clusters according to similarities of the messages; finding portions variable between messages in the messages in each of the clusters; trying to determine a confidential attribute of the variable portions by using predefined rules; and in a case where there is a portion whose confidential attribute is undeterminable by using the rule, determining the confidential attribute of the portion having the undeterminable confidential attribute, by estimating the confidential attribute from another portion having a determined confidential attribute;wherein the step of estimating and determining the confidential attribute includes the step of estimating the confidential attribute with use of a correspondence between an appearance location in the messages of the portion having the undeterminable confidential attribute and an appearance location in the messages of the portion having the determined confidential attribute.