Web 2.0 indicates and proposes new types of internet based service whereas exact definition of Web 2.0 still varies. Blog is one of the representative services of this change. Blog is considered as personal media and, on that account, blog is considered as a fundamental service that can express author’s characteristics.
Nevertheless dramatic increasing number of blog users and distinctive characteristics of a blog service, management and manipulation of a blog page is not easy for common individuals. Therefore, most major companies provide blog hosting service and most of blog user actually relies on such services. However, such blogs are maintained without any considerations about one of the most important features of blog that it represents user’s characteristic. Hence we are suggesting a system that can automatically generate blog page designs with considering each user’s preferences.
Proposed system utilized artificial immune system to provide automatic blog page generation ability. AIS provided a good metaphor in this application and it showed better performance than prevalently used approach, genetic algorithm. With algorithmic performance, it tried to achieve three objectives; fast enough and easy to use, customization chance, trend adaptability. These objectives are accomplished utilizing evolutionary algorithm and designing affinity functions to consider both user’s selection and existing blog pages.
First of all, we defined the ranges of blog design and adapted various techniques to represent blog design as an antibody. Also basic measurements, comparing two different blog designs and quantifying quality of blog design, are designed. Slicing tree layout representation and HSV color space are utilized for blog design representation as well as layout and color comparison methods such as color histogram are adapted for blog design comparisons. Throughout whole process, the system showed ability that can provide new blogs to the user quickly and e...