Link optimization in search engine link analysis

Intermediary transaction http://www.aliyun.com/zixun/aggregation/6858.html ">seo diagnose Taobao guest cloud host technology Hall

In website optimization, it is often mentioned that "content is king, link for the Emperor" now is not only so two points, there are also user experience integrated factors, the content of the king, link for the emperor, plus the user experience is the three-generation search engine of their respective focus, which link analysis is also regarded as the use of the most fire technology, In many black Hat SEO also have many techniques are based on the vulnerability of link analysis to enlarge the use. Chengdu SEO today to summarize the search engine link analysis algorithm. The most important two algorithms in search engine link analysis belong to PageRank and hits algorithm, two are the two models of random walk model and subset propagation model to compute the transfer way of link weight.

First look at the PageRank algorithm, PageRank algorithm is a random walk model developed, PageRank mainly take into account the number of chain and Web page quality two factors to judge the importance of a Web page. However, because PageRank did not take into account the topic relevance of the problem, and then extended the topic-sensitive PageRank algorithm, also known as topic sensitive PageRank, strengthen the theme-related link weight value. At the same time, some people have studied the intelligent walk model and the bias Walk model two link algorithms, these two algorithms make up for the PageRank of the random type, because in normal people browsing the page is not a random click on the page link, but the relevance of the link is more likely to get clicked. Then the extension of the subset propagation model: hits algorithm. Hits algorithm defines two pages, is the hub page and authority page, the Hub page refers to a lot of high-quality Authority page links, such as hao123 is such a page; Authority page is a field and topic-related high-quality pages, similar to the SEO field of seowhy, search engine Baidu, Google. Hits is mutual support, a good hub page must point to a lot of good authority page, a good authority page must have a lot of good hub page point. Hits algorithm problem or more obvious, such as very easy to be malicious use, structural instability, low computational efficiency, and later on the extension of the phits algorithm, here is no more to do the explanation. In view of the advantages and disadvantages of the PageRank algorithm and the hits algorithm, the salsa algorithm is studied, which mainly utilizes the correlation characteristics of the hits algorithm, and uses the random walk model of the PageRank algorithm, which is also one of the most effective algorithms of the link analysis algorithm at present.

Article from Lenny Blog http://xlenny.com/network/700.html