An Approach for Generating and Visualizing Association Rules for Access to News Portal Content
Knowledge discovery, Apriori, FPGrowth, SPADE, data mining, association rules, sequential rules.
This work aims to propose and validate an approach for the generation and visualization of association rules and sequence rules obtained from the content access history data of a Brazilian journal. The proposed approach is composed of four phases: exploratory data analysis (EDA), data preprocessing, generation of association and sequence rules, and visualization of results. The algorithmsAprioriand FP-Growth were used to generate the association rules. To generate sequence rules, the algorithm used was SPADE. Parallel coordinate graphs were used to visualize the association rules and graphs for visualization of sequence rules. An outstanding aspect of the proposed approach is the visualization of the rules obtained using graphic resources to enhance the analysis of the results in support of business decisions and contribute to mapping the users’ access profile. The proposal was validated by using data from user access to a digital news portal.