The selection of the relevant association rules using the ELECTRE method with multiple criteria

Azzeddine Dahbi, siham jabri, youssef balouki, Taoufiq Gadi

Abstract


The extraction of association rules is a very attractive data mining task and the most widespread in the business world and in modern society, trying to obtain the interesting relationship and connection between collections of articles, products or items in high transactional databases. The immense quantity of association rules obtained expresses the main obstacle that a decision maker can handle. Consequently, in order to establish the most interesting association rules, several interestingness measures have been introduced. Currently, there is no optimal measure that can be chosen to judge the selected association rules. To avoid this problem we suggest to apply ELECTRE method one of the multi-criteria decision making, taking into consideration a formal study of measures of interest according to structural properties, and intending to find a good compromise and select the most interesting association rules without eliminating any measures. Experiments conducted on reference data sets show a significant improvement in the performance of the proposed strategy.

Keywords


Data Mining; Associatio Rules Mining; Interestingness Measures; MultiCriteria Decision Analysis; Electre Method; Structural Properties

Full Text:

PDF


DOI: http://doi.org/10.11591/ijai.v9.i4.pp638-645

Refbacks

  • There are currently no refbacks.


View IJAI Stats

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.