Irrelevant Feature and Rule Removal for Structural Associative Classification Using Structure-Preserving Flat Representation - Feature Selection for Data and Pattern Recognition - page 201

Table 10.2 Example of user
session

Session 1:
/
/research.html
/research/topics.html
/research/topics/51-business-intelligence.html
/research/topics/55-e-education-ecosystems.html
/research/seminars.html
/research/seminars/413-presentation-by-eric-feinberg.html
/phd-a-msc.html
/phd-a-msc/scholarships.html
/phd-a-msc/scholarships.html#debii
/about.html
/about/objectives.html
/about/mission-and-vision.html
Session 2:
/
/research.html
/research/centres-and-labs.html
/centres-and-labs/217-anti-spam-research-lab-asrl.html
/centres-and-labs/214-centre-for-stringology-a-applications
-csa-.html
/research/jobs.html
/contact-us.html
Session3:
/
/research.html
/research/publications.html
/research/publications/conf-a-journal-papers.html
/allstaff.html
/allstaff/Research Professors & Fellows.html
/exchange-students.html
/phd-a-msc.html
/phd-a-msc/research-training.html

The integer-indexed tree is then formatted as shown in Table 10.4 . This dataset
format representation was proposed by [ 51 ]. Please note that the second column
( cid ) could be used to refer to a specific entity which the record describes (e.g. User
id). However, in many domains such information is often unavailable, or it has been
intentionally omitted or related through the transaction id ( tid ). Hence, in most of the
tree databases represented in this format, the cid column will simple be a repetition
of the tid column. This is the common format used in the frequent subtree mining
field [ 17 ].