Standard Arpu Calculation Improvement Using Artificial Intelligent Techniques

Recognizing how developing browsing behaviour could result in greater return for service providers through more efficient data usage without compromising Quality of Service (QoS), this paper proposes a new innovative model to describe the distribution and occurrence of behavioural errors in data usage models. We suggest: a) that the statistics of behavioural errors can be described in terms of locomotive inefficiencies, which increases error probability depending on the time elapsed since the last occurrence of an error; b) that the distribution of inter-error intervals can be approximated by power law and the relative number of errors. Comparing immersive similarities of data usage and foraging behaviours according to the Levy-Flight hypothesis, the length of the usage can be feasibly increased with less errors and eventually increase average revenue per user (ARPU). The validity of the concept is demonstrated with the aid of experimental data obtained from test software called Learn-2-Fly which sought to make browsing behaviours more efficient through user responses to stimuli created by an artificially intelligent engine. Although there were limitations on the scope of this test, a noticeable change in the user browse duration occurred over the duration of testing periods, with test subjects spending more time browsing and reacting to intended visual stimuli. The study establishes the opportunity to provide a higher quality of service to the end-user, whilst also offering a dynamic opportunity to increase revenue streams. Further consequences, refinements, and future works of the model are described in the body of the paper