Optimization Techniques to Speed Up the Page Loading in Custom Web Browsers

Abstract

Web automation applications are widely used for different purposes such as B2B integration, web mashups, automated testing of web applications, Internet metasearch or technology and business watch. One crucial part in intensive web automation applications, that require real time responses, is for them to execute the navigation sequences in the shortest possible time. The approach of building the automatic web navigation component by using the APIs of conventional browsers, followed by most of the current systems, is not appropriate in that scenario, because it presents performance problems. Other approach consist in creating custom browsers specially designed for web automation, which can develop some improvements based in the peculiarities of the web automation tasks. In this work, we present a new set of techniques and algorithms that allow the parallel evaluation of the scripting code when a custom browser loads a web page. We also outline the components that should be included in the custom browser architecture to implement these techniques. The tests executed with real web sources, to evaluate the validity of our proposal, show that a custom web browser loads the web pages faster when the scripts are executed in parallel using the designed techniques.