【作者】 王丹丹；

【导师】 路松峰；

【作者基本信息】 华中科技大学 ， 计算机应用技术， 2007， 硕士

【摘要】 随着移动通信的发展,人们迫切希望能方便地利用手持移动设备访问Web网站。由于移动设备的小屏幕和低带宽的缺点,使得这一难题一直没有得到很好的解决。Web网页是为桌面计算机而设计的,桌面计算机的屏幕大,翻页和滚屏十分方便。因而Web页面的内容很丰富。但移动设备,例如手机,具有内存小、上网带宽低和屏幕小的先天缺点,不可能把一个完整的Web页面直接映射到移动设备屏幕上,因而如何在小屏幕设备上方便地浏览Web网页一直是研究的热点。分析在Web信息分块领域中已有的研究工作,其中包括纯HTML文档的转换以及自然语言的总结等等。大部分算法都是单纯的利用HTML标签之间的关系来对Web文档进行解析,这些分块算法对固定格式以及规范标签用法的页面来说,分块效率以及正确率会比较高,但是,现在的大部分Web页面都是复杂,不规则以及含有太多噪声的,例如导航链接以及广告链接等等,这时这样的分块算法的效果就不是很理想,在此基础上设计出两种适合于移动设备小屏幕的Web页面分块算法,一种是基于DOM TREE的页面分块算法,将DOM TREE的叶子结点看做是不可分割的最小语义块,然后再根据结点之间的语义关系进行迭代,通过语义块的提取,分割条的检测以及权重的设置,找到各个页面分块,再根据内容结构的重构形成适合于手持终端显示的页面。另一种算法利用Web网页上对象的位置信息对信息块进行逐层聚类,生成一棵网页分块树,再根据移动设备屏幕的特点把网页分块树转换成适合小屏幕浏览的页面。设计并实现了面向移动设备的Web页面分块算法的小型实验系统,系统能有效地将Web页面转换成适合小屏幕浏览的页面。更多还原

【Abstract】 With the rapid development of telecommunication, people want to access Web pages by using mobile phones just like they use PC, but the problem has been not well solved due to small screen and less computational power of the mobile device.Web pages are originally designed for browsers of personal the computer which has a big screen and is easy for paging and scrolling. Since the content of Web pages is rich while mobile device has very little screen and memory to support the whole pages at once, the conversion from Web content to content that suit mobile device to display has become a popular field.The paper first analyses recent research on this field, including pure html document conversion and natural languages. Most of the conversion algorithms are based on the relationship of html tags, but html tags are not very accurate to categorize the content of the Web pages into blocks, each of which to be displayed as a single page on the screen of mobile device. So new algorithms are proposed to solve this problem, one is base on Dom Tree and visual features of each Web page. Each leaf node of Dom tree is regarded as a smallest semantic block. Through semantic block extraction and segmentation detection and weight value, Web page blocks are found and reconstructed so that they are suitable for mobile device to display. The other first clusters information blocks (which composes the Web page) hierarchically into a block tree. And according to the features of the screen of devices, it then converts the nodes of the tree into pages suited for small screen.On the base of the research, a small prototype is developed to change the Web page into small screen pages.更多还原