An Efficient Prediction Based on Web User Simulation Approach Using Modified Ant Optimization Model and Hierarchical Clustering

Ant colony algorithms (ACA) are to solve difficult optimization problems, such as the traveling salesman, and have since been extended to solve many discrete optimization problems. ACA are derived from the process by which ant colonies find the shortest route. Here an ant colony optimization based algorithm to predict web usage patterns is presented. Our methodology incorporates content, structure as well as web usage data. Ants learn from the clustered real Web user data subsequently, trained ants are released onto a new web graph and the new artificial sessions are compared with real sessions, previously captured via web log processing. The main results of this work are related to an effective prediction of the aggregated patterns of real usage, which reaching near about 87%. Moreover, this approach obtains a quantitative representation of the keywords from the content data that influence the sessions. The proposed work and innovative research is on the basis of the improved next node election, through which we get an improvement in basis ant learning behavior algorithm. This modified ant behavior learning algorithm predicts a larger matching sequence size of real website user sessions along with increased learning rate of the software agents that means most of the ants reach to the uppermost threshold most of the time which directly turns into increased prediction correct rate.