Nowadays, there are many new web sites occur daily on the Internet and the demand for web page creation and design increases. The problem is it takes time to create and design the web pages. Moreover, designing web page templates to satisfy user preference is difficult. In order to reduce time and allow users to interact with the design process, we propose the web application which applies interactive genetic algorithm to generate web page templates and allow users to get involved with the system.
The proposed algorithm evolves HTML files for the content and CSS style sheets for a presentation format. Users can rate each section of a web page template according to their preferences. The given scores feed back to the system and the generated template is evolved. The experiments are conducted by assuming user’s preferential criteria. We designed score-rules to test the algorithm whether it can create web page templates to satisfy the criteria or not. The experimental results show that the proposed algorithm can generate web page templates that meet the criteria.
Web sites are very popular at present. Many new web sites occur daily on the Internet ; thus, the requirements to create and design web pages are increased. These require users to know programming languages and tools for developing and designing web pages. It takes time to develop and design web pages as their need. In order to reduce time from both creating and designing web pages, they have to find a new solution by using web page templates. Web page templates consist of HTML files and CSS files. HTML files hold the structure and content of the page and CSS Style Sheet files hold the presentation styles of pages.
Therefore, we propose a new web application which is developed by applying interactive genetic algorithm which requires the involvement of users to interact with the web application. The interactive genetic algorithm has been applied to many applications in design such as a fashion design which is used to model women’s dress , a Japanese Kimono design to model Yukata which is a traditional Japanese garment often worn in mid-summer , a font generation system which is designed to emerge various fonts based on user’s Kansei without hand drawing , evolving colors in user interfaces to search for a solution that provides a good trade-off between aesthetics and accessibility requirements , an office layout support system which can generate not only in square space but also in polygonal space , a user interface design which evolve user interfaces in the XUL interface definition language which is a user interface markup language developed by the Mozilla project , a sign sound design which is to generate melody based sounds freely and easily , a web site design system which has users to be involved in the process to generate web page [9, 12].
The difference between this work and the previous work [9, 12] is how to rate a web page template. In the previous work, users are involved in the evaluation process and they can rate a preferable web page template one at a time. The problem is we cannot know which section of the web page template they like or dislike. Thus, we have divided the web page template into many sections and users can rate each section of the web page template for what they like or dislike. By applying interactive genetic algorithm, the proposed algorithm generates many web page templates with different layouts for users to choose. Users just rate each section of the web page templates, such as header, footer, sidebar, etc., to express their preference how much they like or dislike it.
Then, the algorithm will evolve a new population of web page templates based on the old ones according to users’ preferences Users can view and download web page templates which are created by the web application to apply these templates to their creation and design tasks. The details will describe in the remaining of this paper which is organized as follows: section II describes the background of genetic algorithm and interactive genetic algorithm, section III presents the proposed method, section IV describes the experiments and results, and section V is a conclusion.