This is my website slicing algorithm on the idea of a summary. 

  Before I wrote: A Web-oriented search engine block, slicing the principle of achieving and demonstrating, as the in-depth work, and gradually meet the following questions: 

  Slice the page size: 

  Web slicing algorithm purpose is not to find the precise needs of the content, but rather recognition of the various functions of the regional website, navigation, links, content, advertising and footer areas, and other areas. 

  Slice the target Web page: 

  Internet website features yarn about two kinds of types, type and content directory; with the development of search engines and site structure to a flat gradually in the direction of development of this car East has also made data validation, and along with Display resolution of the continuous enhancement of a combination of content and directory website showed an increasing trend, involving天极website can be said to be one of the examples. 

  Web slicing algorithm should be the object of target: content type and content directory mixed.    On different pages, there should be a recognition algorithm, which should include standards? 

  The largest scope of the content of the sites identified: 

  From the size of the slice can be seen, it should be the content areas as a part of a separate cut out.    Under the general law of web design, there are two kinds of general contents of the house: 1, including type (such as blog) 2, tie type (such as bbs posts). 

  If the deal with the content of Web page: 

  Now, most sites to improve the user experience and increase the number of pages demonstrate the need to do a web page address, in part by the need to set up another. 

  Inadvertently see: VIPS: Based on the visual Web pages paging algorithm, theoretically proved the feasibility of this approach.    But there are many obstacles to realize, as the Honourable said: 

  Snails published in 2006-02-21 12:40 AM IP: 220.184.129 .* 

  I used the other day by floating location positioning instead of absolute positioning, and client JavaScript dynamic array.    Client with the object of dynamically generated script insertion. 

  Longsi him.    See how his analysis. 

  This algorithm is too dependent on achieving a specific, it is difficult to have a good solution. 

  Moreover, now rely on the client-side script dynamic performance of the show slowly beginning of the epidemic, this algorithm to adapt to the wave of the future. 

  Take the most simple, I have a similar tool OutLook style of the pages are generated by the script, I see how his analysis! 

  Visual analysis can only be adhering to the visual, only the static pages on the screen for the correct analysis can be divided horizontally and vertically and horizontally and vertically divided simple algorithm is easy to be done, but to sum up, these elements of the division to the Block is difficult. 

  There is only one good way to simulate the keyboard mouse, keyboard response to the return of the object, which in IE can be achieved.    This segmentation can be achieved good bar ownership of the objects. 

  How I would like to think that I have a simple algorithm than in the text of the algorithm is much easier. 

  Visual segmentation bar on the screen, a very simple blank for the expansion - narrowing the algorithm, this left blank will gradually Clear out words blurred.    Then a fuzzy treatment, and then use a picture brightness threshold to convert binary map, and then make a vector processing, leaving lines.    Attributed to 0 degrees and 90 degrees forward, get a fragmentation of the vector. 

  Then, in accordance with a density of each mouse click access to simulate object! So that it could be summed up by the latter. 

  Why is necessary to analyze the HTML? Very many cases, but to fundamental analysis. 

  I am the current progress is: can identify navigation area, links, footer area. 

  Analysis of the content is a difficult, in the light of my own needs, as long as the content of the largest found on the district. 

  This time of the algorithm is an algorithm to solve specific problems.    The textbooks are the most general algorithm commonly used in a description of the method. 

  To use the saying goes, to solve practical problems is a good method of approach.    But can we solve the problem expressed by mathematical model is still a level algorithm continuously improve the Basic requirements.