DOM-based Web Pages to Determine the Structure of the Similarity Algorithm

Web data is currently mainly in the form of HTML pages, expressed by the HTML language of Web pages through the browser after analysis is only suitable for people to browse, not suitable for data exchange as a way to deal with by a computer. This article will make web page decompound a DOM tree, then from the DOM tree body root node to start, in accordance with the breadth-first traversal order DOM tree, layer by layer comparison DOM node tree, statistics of its changes, and then the sum of all floors of the changes, If less than a certain threshold, it is structurally similar to two pages, otherwise dissimilar, because this algorithm is only concerned about the page structure information without concern for the content of the page, it has a very high operating efficiency, while the algorithm is not limited to a specific web page, with good versatility.