Three steps to perfect build link platform data collection analysis

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

We know that data collection is a coolie live, the need is tenacious perseverance and efficient execution, data analysis is a mental life, need a meticulous heart and all-round thinking, the ultimate goal is to achieve the data to speak. In the final analysis is a relationship between thinking and implementation, we have to straighten out our thinking, with strong perseverance and efficient execution to create a perfect data collection and analytical work.

Receive this collection analysis Friendship chain platform task, we are in the "three steps", first thought should be how to analyze a link platform of value evaluation criteria? And then how to find these link platforms? Finally, how to effectively and better organize the data?

The first step, determine the standard of Friendship link platform

Here we must first think about, if we are an ordinary user, then I want to link the platform for what? One is to do friendship links, the other is pure to do outside the chain. So the following seven major standards.

1. Alexa rankings. Alexa ranked the inevitable popularity of high.

2. Baidu Google included. Many of the linked platform does not prove that the total number of links to this platform, but must be able to prove that the Web site published links outside the chain of strong effect.

3. Yahoo. Outside the chain. Yahoo's outside the chain of evaluation in the SEO industry is generally accepted more reliable.

4. Baidu Snapshot. Daily snapshots show that the site is still a good weight for Baidu.

5. Daily update. An important reference standard for active degree.

6. Domain name registration time. Now the search engine on the old domain name is relatively high weight.

7. Total number of links. Can prove that this site is rich in resources, on this platform can generally find the link you want.

The second step, find links platform.

1, the use of keyword search. The general Link link platform will be such a series of key words "friendship link", "Link Platform", "Link Platform" "Exchange links", "Friendship Link Exchange", "Friendship Link Transaction" and a series of related words. These in Baidu's Drop-down box or search engine related search can be seen, really can't think of friends can also find a link platform to see what the keyword is, and then based on these keywords to search.

2, the use of a good site search platform. In some platform sites such as Portal blog, well-known webmaster forum and Tencent QQ Group search "Friendship link platform" can find the relevant link platform. Why this is so, because any one want to put the link platform to do the station will not let go of these propaganda stage, this time you search in the station can find them. Of course, in the relevant platform to find better results, such as this task Wu Jian students in Go9go this well-known link platform to find more than 40 sites required.

3. Use search engine syntax "Inurl". We found that a large number of linked sites are used by the same template, its exchange of linked files are generally exchange.asp or exchange.php, then we can use the "Inurl:exchange link" (Note the middle of a space), This syntax means that search engines will search Web pages that contain the words "exchange" and "link" in the Web site, and then add the word "link" after "Exchange" to make the search results more accurate.

The third step, the concrete collation data work.

1. Avoid duplication of data. There are students to use the Favorites method, the Web site one by one open still buckle collection, but the browser's favorites are not allowed with the domain name of the Web site exists, so the collection of URLs is unique. The name I recommend here is the Advanced filtering feature in Excle. Put all the collected URLs into excle (add only after other list of URLs), then select all the data, click "Data"-> "filter"-> "Advanced Filter" will "Select Not duplicate Record", then click OK button, OK duplicate record not.

2. Carefully organize the basic data. Here in particular to put forward is "meticulous" two words, we do data work is to use data to speak, the data are wrong then we are saying is untrue. Like my "Google included in the list" inside the Go9go data mistaken, got the MU teacher criticism. So for this type of data I have to carefully fill in the data, at least two times to determine. When we encounter some abnormal data, we should make a special mark. such as this address http://tool.haiphp.com/Baidu included unexpectedly nearly 36W, here I will be particularly marked, and then check the reason.

3. Refine the underlying data to classify it. In front of some basic data sorted out, but not intuitive can not give people a clear feeling. Here I'm going to refine the underlying data. For example, Baidu/Google ranked Top10,alexa ranking Top10, the key words Baidu/Google ranked Top10 and so on. This step is to use the data to talk about the part, of course, how to subdivide here depends on your thinking.

Original article please indicate reproduced from: Sem said

This article address: http://www.semsay.com/essence/27.html