Models' Specifications to Build Adaptative Menus

Web engineering becomes increasingly important in the last years. WEB and audio services have to provide the best services possible. To achieve this goal, they have to find out what the customers are doing without altering their privacy. This paper presents two classes of models, mathematical and learning models, and two possible ways to manage and build adaptative menus. These methods are gravity analogy, learning by sanction reinforcement. Later on, a comparison of these two models will be made based on two criteria: efficiency (answering time and computer load) and accuracy with customer expectation. The final step will be to carry out psychological analysis of user activity, meaning, “what is my perception of time into and between service consultation” to determine ways to set parameters of such a system.