Master Thesis

Web page loading time is one of the critical challenges on the Internet. Thistopic attracts more researchers as it largely reflects the user satisfaction.A web page resources typically consists of Hyper Text Markup Language(HTML), Cascading Style-Sheet (CSS), JavaScript (JS), media files, etc.As the complexity of the web page increases, the size of files fetched fromserver also increases which causes delay in transmitting, loading, parsing andrendering the web page.The web page is not rendered until HTML, CSS and JS files are loadedand parsed. Not all such files are needed to be rendered for the first page.Optimizing and loading only critical les required for the first page, can considerablyreduce time taken to load the web page. After the first page isrendered, the rest of the files can be loaded and executed. In this thesis,we focus on improving the time taken for First Meaningful Paint(FMP) ofwebsite by loading the critical files initially. We used Design Science Research(DSR) methodology and created a PhantomJS application, to identifyand segregate used and unused CSS for the first meaningful paint of the webpage. We also analyzed script files to load only critical files required forFMP and the rest later. We evaluated our experiment by studying a case onAxis Communications AB and measured its website's performance to find ifour experiment reduced FMP time. It showed improvement in time takenfor FMP from 1284ms to 372.6ms. Based on the results, we suggest to webdevelopers to separate and load the files required for FMP and load the restof the files needed for the website later, so that the first page is renderedquickly