Applying Triadic FCA in Studying Web Usage Behaviors

Formal Concept Analysis (FCA) is well known for its features addressing Knowledge Processing and Knowledge Representation as well as offering a reasoning support for understanding the structure of large collections of information and knowledge. This paper aims to introduce a triadic approach to the study of web usage behavior. User dynamics is captured in logs, containing a large variety of data. These logs are then studied using Triadic FCA, the knowledge content being expressed as a collection of triconcepts. Temporal aspects of web usage behavior are considered as conditions in tricontexts, being then expressed as modi in triconcepts. The gained knowledge is then visualized using CIRCOS, a software package for visualizing data and information in a circular layout. This circular layout emphasizes patterns of user dynamics.