Millions of web pages on the world wide web provide us an edgeto understand the design structure, practice and building blocks on largescale. Every page consists of solid example of visual problem solving andaesthetics. Data mining, machine learning, knowledge discovery and cloud haverevolutionized the web, crawling procedures, advertising platforms, agent systemsand recommender systems being used by thousands of people day to day. But,previous data mining techniques only focus on content extraction without theactual presentation of contents.
This report provides a theoretical frameworkthat focuses on how customized design mining done from web using differenttools and techniques for attractive and flexible website and mobile applicationdesigns.Key wordsDesign Mining, Design Extraction, Web Content Mining, Webmining, Graph miningIntroductionThe Web has transformed the nature of creative work. For thefirst time, millions of people have a direct outlet for sharing their creationswith the world. As a result, the Web has become the largest repository ofdesign knowledge in human history, and the ensuing democratization of designhas created a critical feedback loop, engendering a new culture of reuse andremixing. The means and methods designers use to employ to draw on priorwork, however, remain mostly informal and ad hoc.
How can content producersfind relevant examples amongst hundreds of millions of possibilities andleverage existing design practice to inform and improve their creations? Designs are always derived from data. Whether building a simpleWeb page, programming an interface of any application, or planning the menu forhandheld devices, designers draw on prior work to solve new problems. Bothnovices and experts alike turn to examples of previous work to reduce barriersto entry, lower the cognitive burden of routine tasks, and foster inspirationwhen innovation is required. In fact, using examples to explore designvariations is crux of creativity. The Web provides an opportunity to use datato inform design practice on an unprecedented scale. The ready availability ofcreative work online has engendered a new culture of remixing, in whichcontent producers in diverse fields routinely leverage, adapt, and repurposeexisting artifacts in their own creations. For Web design in particular, eachof the billions of pages on the Web today comprises a concrete example ofvisual problem solving, creativity, and aesthetic preference, all in a formthat can be easily accessed and shared. The means and methods Web designersemploy to draw on prior work, however, are largely informal and ad hoc.
Whilesearch engines like Google, MSN etc. have revolutionized the process oflocating information on the Web, no such broad support exists for findingrelevant designs amongst hundreds of millions of extant pages.Similarly, while end-users can “view source” to inspect apage’s implementation, they still shoulder most of the burden of parsing apage’s code to understand its design, and manually adapting that design to thetask at hand. Without better tools to help users locate, understand, andleverage existing work, much design data on the Web remains underutilized. Thisreport introduces design mining for the Web using data mining and knowledge discovery to catalog,analyze, and adapt the design of Web pages. It demonstrates how many of thesame technologies that drive search engines, advertising platforms, andrecommender systems on the Web today can be repurposed to support new, usefuldesign interactions. By focusing not on the information content of Web pagesbut the way that content is presented, design mining enables the development of data-driven designtools and rigorous statistical analysis of Web design patterns and trends.