Absrtact: The preface this article is suitable with the large-scale website SEO personnel, the small website may also refer. The purpose of this paper is to explore the content potential of the website, to present the content that the users may care about, to satisfy their needs, and to obtain the corresponding SEO flow.
I. Preface
This article is more suitable with the large web site SEO personnel, small sites can also refer to.
The purpose of this paper is to explore the content potential of the website, to present the content that the users may care about, to satisfy their needs and to obtain the corresponding SEO traffic.
Many large web sites are using a method, but few people come out detail explanation.
This kind of SEO traffic is how to obtain, below give a plain example.
Let's say I search for "ios Hero 3" on Baidu to find a Hero 3 game that can run under iOS.
In fact, the game doesn't exist. So there is no single page that allows me to find what I want (if there is also the title Party)
So I went to Tgbus's page about the iOS Hero 3.
On this page I found the Hero 2 game on iOS and other iOS games like Hero 3, and some weird news (OK, this page isn't very easy to read)
Finall, I downloaded the 2 iOS version of the hero in Tgbus to try it.
OK, let's change the real life example:
A girl went to buy clothes, so she saw a pink dress, but not her size.
What does a salesperson do, yes, recommend a dress that resembles a color or style and has a code.
(Tgbus seems to recommend a pair of jeans.) )
So, how do we find out what the user wants to dress and the dress in our warehouse, and at the most appropriate time to give the user the best results, but also to obtain the SEO traffic, this is not a very simple thing.
Wait, isn't this just a recommendation engine? This is a complicated thing for an engineer to do. In fact, most of the time, this is only SEO wishful thinking, engineers will not come to bird you what recommendation engine, we first from an executable angle, self-reliance to carry out this SEO method.
Two. Content analysis, keyword analysis, data interface design
A SEO know how much content of their site is very important, often encountered and people say: "Your site this xx page has a problem" ah? What page is this, I've never seen. ”
General a website perpendicular to the score, there is a homepage, content page, List page.
Content pages, and may be divided into picture pages, comment pages, article paging, and so on
In the list page, it is possible to divide channel pages, product list pages, index pages, topic pages, and so on.
A general set of pages corresponding to a set of even multiple sets of PHP templates.
Need to figure out whether these templates are in a schema, public database, which fields are used on the page, it is best to find the developer of the corresponding template, if conditional request to the source code to view permissions, you can see for yourself. A content aggregation of requirements can be achieved to a large extent depends on these content, first clear the ingredients are complete, and then start cooking, otherwise bricks.
Cross-sectional may be more to business direction, such as we have to sell lines, selling tickets, selling hotels, Raiders, user pictures, forum posts and so on each channel, each channel may be by different departments in charge. Which is the website hot, need to push (at least you do a SEO things have commercial value, in the electric business Company is very important), including whether the channel is still operating, a perennial unattended channel, is clearly not a good source of content. In general, the main product line, and UGC content is generally the content of the site to provide the main force. If we want to tap into user needs, we can also prioritize them.
This step takes a lot of time, and complex sites take 1 months to figure out exactly how many types of pages the site has. After figuring out the problem,
Next, you need to know just how many types of content there are.
For example how many sku, how many articles, how many posts, how many tags, how many categories and so on
This is a lot of people who think poorly of doing similar work and do it by feeling. Finally make a lot of duplicate content of the page, repeat the page on the SEO has how bad impact will not have to mention.
The method of counting the number of contents (from good to bad sort).
1. Read the database
2. Some "ingenious" ways to Count
3. Use tool to catch
4. Guess by experience (basically not reliable)
Reading a database is the simplest and most accurate way to do it, a select finished
If you don't have database permissions, you need to find out. For example, how many articles, that can calculate the number of pagination * per page of the number of articles to statistics
If it is the ID, then to gather the number of self-increasing ID;
If the data is fixed format, such as the pictures to Beijing, the weather in Beijing, the number of regions * type to calculate, and so on;
Through the tools to catch is a lot of SEO dream, countless people asked such a question, there is no tool to count my site how many pages ah.
Sorry, really did not, because of a variety of web site reasons, no one tool can be counted on a large site in the end how many pages (who have that tool than Google, Baidu Crawler also NB), too many reptile traps, blocking the deep crawl of things. Of course, such a crawl tool is not useless, for small sites, or specific channels, or even specific blocks of the crawl, still have some effect. For example, Httrack,xeun, of course, I prefer to use the locomotive. Scripting languages such as Python,shell have always been omnipotent.
The above is basically content analysis of the general situation, to find out the type of content, quantity and operation, SEO is a lot of benefits.
Three. Keywords mining, cleaning and filtering
The key word is difficult to dig, say simple also not simple.
Basic everyone will ask, how to mining keyword production thesaurus. Let's talk about some common techniques.
1.baidu/google API
2. Collect Love station, Chinaz, Bole and other data
3. Collect Baidu dropdown box (other search engine empathy)
4. Collect Baidu related search (similar to other search engines)
5. Site search and Natural flow keywords
6. Ready-made dictionary/Thesaurus
Some of the points of attention in the concrete realization are the things that are summed up in the process of practice. Method said, basically 10 people inside 1 people to practice some of the good
1. Baidu and Google API is to apply, to find a way to get a, if not, can only use Baidu web-level collection, Baidu auction background often changed, so is not very stable, this side of the recommendation of a tool http://www.lingdonge.com/(temporary record may not open , the author is very NB, engaged in the Knight Station group software. Baidu API Python SOAP communication has bugs, not even (perhaps I level too low t_t), PHP will soap communication words can write their own script to run, Google did not play, should be almost;
2.API is the data with accurate search quantity, so it is the first data source of thesaurus;
3. Baidu Drop-down Box Reverse collection do relatively little, the collection address for http://suggestion.baidu.com/su?wd=xxxxxxxxx+ a pile of parameters, specific adjustments, a small script can be done, but the data depth is limited, the general collection of 2 rounds will not have to pick up , the basic production of new data;
4. Related search can use the locomotive or fly to the golden flower such as, because it is collecting serp, reverse collection of this piece to find ways to bypass;
5. In-site search and natural flow keywords to GA to do it on the line, batch export do not say. Very simply, not using GA may be tragic some;
6. Pinyin Input Method Thesaurus, recommend a data site http://www.datatang.com/, you can see, there will be some industry thesaurus;
7. Some special search engines, Taobao, Youku and the like, they also accumulated a lot of data information.
Four. Word/search/Sort/fix
Five. Channel operation, maintenance, development
Six. Data monitoring