Access classification device, access classification method, and recording medium

An access classification method executed by an access classification device that classifies an input access into a set with similar features, the access classification method comprising:

• constructing a plurality of trees in each of which at least a first destination and a second destination are set as nodes, content information corresponding to the nodes is added to the nodes, and an instruction to transfer an access from the first destination to the second destination is set as an edge;
• associating nodes of the plurality of trees with each other for the plurality of trees constructed at the constructing, based on similarity between local structures of the trees;
• calculating similarity between the nodes associated with each other in the plurality of trees, based on the content information added to the nodes, and calculating similarity between the plurality of trees using the calculated similarity between the nodes associated with each other, by a processor;
• and
• classifying the access into a set with similar features, based on the similarity calculated at the calculating.

A non-transitory computer-readable recording medium storing therein an access classification program that causes a computer to execute a process comprising:

• constructing a plurality of trees in each of which at least a first destination and a second destination are set as nodes, content information corresponding to the nodes is added to the nodes, and an instruction to transfer an access from the first destination to the second destination is set as an edge;
• associating nodes of the plurality of trees with each other for the plurality of trees constructed at the constructing, based on similarity between local structures of the trees;
• calculating similarity between the nodes associated with each other in the plurality of trees, based on the content information added to the nodes, and calculating similarity between the plurality of trees using the calculated similarity between the nodes associated with each other;
• and
• classifying the access into a set with similar features, based on the similarity calculated at the calculating.