基于视觉特征的网页最优分割算法

Web Page Optimal Segmentation Algorithm Based on Visual Features

网页分割技术是实现网页自适应呈现的关键。针对经典的基于视觉的网页分割算法VIPS(Vision-based Page Segmentation Algorithm)分割过碎和半自动的问题,基于图最优划分思想提出了一种新颖的基于视觉的网页最优分割算法VWOS(Vision-based Web Optimal Segmentation)。考虑到视觉特征和网页结构,将网页构造为加权无向连通图,网页分割转化为图的最优划分,基于Kruskal算法并结合网页分割的过程,设计网页分割算法VWOS。实验证明,与VIPS相比,采用VWOS算法分割网页的语义完整性更好,且不需要人工参与。

The Web page segmentation technique is a key point to realize Web page adaptive presentatiorL To overcome the shortcomings of the classical Web page segmentation algorithm VIPS（Vision-based Page Segmentation Algorithm） including fragmented content and semi-automatic, a novel Web page segmentation VWOS（Vision-based Web Optimal Segmentation） was proposed based on the optimal division of graph. The Web page is constructed as the weighted undi- rected connected graph from the perspective of visual features and structure of the Web page. Therefore, the problem of Web page segmentation is transformed into the optimal division of graph. VWOS was designed by combining Kruskal algorithm and the process of the Web page segmentation. It was proved by the experimentation that the effect of Web page segmentation produced by VWOS is better than that by VIPS.