SEO Deep Dive: Content Breakdown with Natural Language API

You for sure heard it: “Content is King”. One of the most quoted statements when it comes to SEO. I’d like to take things even further and show you one (easy) method on how to validate your content. The key to this method is the Google Natural Language API.

The Content Breakdown in Theory

Let’s assume, you have a text. You read it and you understand it. Maybe you are reading it twice to get a clearer picture regarding its structure, keyword usage and so on. The more you are used to SEO the more you might tend to have a first feeling if a text is well written for your SEO goals. But it still might happen, that the Search Engines do not understand your content the way you would like them to. Today a text is written for the user, no longer for any Search Engines or Algorithms. But still, they are machines, nothing else, even so, they begin to understand human behavior better and better. The only and in my opinion the best way to check that out, is to check your text with Googles Natural Language API. But why is that? Why can that tool help you any better than other tools?

Understanding the Structure of written content

Written content is still one of the most important information on a page. Of course not only for the Search Engines but also for the user. All Search Engines are trying to understand more of the user behavior and machine learning is one way to get closer to understand the human way of thinking. That’s the reason why SEO specialists should learn to understand “entities”, “sentiments”, “syntax” and “categories”. But it’s important to not only understand these terms but also to understand why Search Engines are “thinking” that way. It’s more or like all about “scoring” texts based on different parameters. And the Deep Dive behind that is to understand how Search Engines are doing that. In the end, this is a requirement for the improvement of voice based search results. By having a look at Googles Natural Language API, you can see and begin to understand how Google understands written content on a website.

The elements explained in short

Entities

With the Hummingbird Update on 26th, September 2013, Google introduced entity search into their search algorithm. Entities are all about understanding the users search intent while Google is mapping additional sources with other verified sources that might answer questions.

Sentiments

The sentiments analysis tries to find out more about the intent of a text. Google tries to find an answer whether it’s a objective or subjective text. It also tries to find out whether it’s a positive, neutral or negative text.

Syntax

While checking the syntax of a text, Google tries to understand more about sentence length, sentence structure and so on. This syntax check helps Google to understand more about a possible reading level of a text.

Categories

By “scanning” a text, Google also tries to find out which category a text can be assigned to. This check could result in the classification of a text as news article, if it’s about computers and electronics, etc..

Detailed information on Natural Language API

Less than 5.000 API requests are free (more information about the pricing can be found here) while 1.000 characters count as 1 API request. Anyone with a Google Account may use the Natural Language API This solution is way cheaper than other solutions out there

How to use Googles Natural Language API

Step 1: Login or Sign Up to Google Cloud Services

Sign yourself in here by using the “try” button on this website. In case you already do have a cloud account, you can directly login on the same website. In case you are new to Google Cloud Services you’ll have to give Google your payment information. In return you get a $300 Credit which can be used within one year after signup.

Step 2: The Cloud Platform Console Screen

After either signing in or signing up, you will arrive at the Google Cloud Platform Console screen. In this screen use the search bar on top and type in “language”. There should be just one language API entry in the search results. Choose “Cloud Natural Language API”. You’ll be forwarded to the API landing page. Enable the API now. This might take a little while.

Step 3: Billing and Request Information Screen

As soon as the API is enabled, you should see a screen, which shows you some billing and request information.

Step 4: Using premade functionality

Now you are able to use the premade functionality Google already included in your account. You can find the links to the request forms here. The rest is self-explanatory. Choose one of the API-Requests offered by Google and click on the link. A new screen opens where you can choose on the one hand which fields you’d like to view in the final result. The other form field currently just has to brackets {} included. Please make sure that your setup looks something like this in case you’d like to copy past text from a document or URL.

{
  "document":{
    "type":"PLAIN_TEXT",
    "language": "EN",
    "content":"'Lawrence of Arabia' is a highly rated film biography about \
                British Lieutenant T. E. Lawrence. Peter O'Toole plays \
                Lawrence in the film."
  },
  "encodingType":"UTF8"
}

Try the Google Cloud Natural Language API now

Take the following text:

An entity is something that exists as itself, as a subject or as an object, actually or potentially, concretely or abstractly, physically or not. It need not be of material existence. In particular, abstractions and legal fictions are usually regarded as entities. In general, there is also no presumption that an entity is animate, or present. The word is abstract in intention. It may refer, for example, to Bucephalus, the horse of Alexander; to a stone; to a cardinal number; to a language; or to ghosts or other spirits. The word entitative is the adjective form of the noun entity. Something that is entitative is considered in its own right.

And copy it in the API fields with the following settings: When you are done, click either on “Authorize and Execute” or on “Execute without OAuth”. Both are fine and should work. As a result, you get an API-Output with all the available data. Now, copy the output below the query and go to my json to xml converter. Paste the JASON Output into the form and click on “Convert JSON to XML”. Now copy and paste the output and save it in a plain text document on your local computer with .xml file suffix. Now you can open that XML-File in Excel and sort your table as you like.

Other Natural Language API providers

• textrazor
• IBM Watson