1. What is NLP?
NLP stands for Natural Language Processing, and it is the ability of a computer program to understand human language as it is spoken and written.
Initially, NLP was introduced by Google during the BERT algorithm update. They created a bunch of new metrics to improve the interpretation of search queries and to understand what makes better content. It recognizes the user's intent and the overall sentiment. By sentiment, we are talking about overall emotion for a word, sentence, and page as a whole.
But something was missing as the update did not really aim at helping marketers to generate quality content to the fullest. We've noticed room for improvement, and we decided to create our own NLP engine - Surfer NLP.
While Google created its NLP algorithms for general purposes, Surfer NLP is a little bit more SEO oriented. It will help you optimize your new articles from the beginning or enhance your old, existing content. Each top Google search result is measured and analyzed with AI-powered techniques making SEO efforts more effective. Since search engines use machine learning technology, so do we!
2. How is NLP used at Surfer?
NLP is available for SERP Analyzer, Content Editor, and Audit. You don't have to turn it on manually for your queries. All you need to do is pick a location associated with one of the supported languages listed in the next section of this article, and your query will automatically be enhanced with NLP.
In Audit and SERP Analyzer, you can additionally enhance your query with NLP Sentiment. This option, however, has to be manually enabled.
Depending on the language of the query, Surfer uses either the NLP API of Google or our own NLP engine to extract data about entities and sentiment. To deepen our analysis and make the data actionable, we cross it with our True Density calculation. Regardless of which group your language is in, we will support you in producing great content and help your page rank high.
3. Which languages are supported?
We’ll gradually add the most frequently requested languages, but right now, the Surfer NLP is available in 4 languages:
English, Danish, Dutch, Polish
As the NLP features run in connection with Google's NLP API, we are able to support the languages that the API supports, which are:
English, French, German, Italian, Japanese, Korean, Portuguese, Russian, Spanish, Simplified Chinese, Traditional Chinese
4. NLP in SERP Analyzer
NLP entities have been incorporated into the SERP Analyzer. You’ll find them in the True Density section, where relevant phrases and words that have been labeled as NLP entities are marked with the NLP icon.
Additionally, you can decide if you would also want to see NLP sentiment section in your report. In terms of Sentiment, we use similar procedures to IBM Watson and measure the overall sentiment of page content.
If you're interested in finding out more about our decision to use IBM Watson for page sentiment, Slawek explains it in our article on NLP and on-page SEO.
This is how you can access NLP Sentiment on SERP Analyzer:
1. Make sure you opted to see the sentiment by turning on the toggle when making a query.
2. Go to the list of factors on the left and check the 'Sentiment' box.
The sentiment score ranges from 1 to -1. The higher or lower the score, the more strongly that negative or positive sentiment is expressed.
The general rule is it would be more challenging to rank a page that has negative sentiment if the top-ranking pages all have a positive sentiment. This is most likely how Google understands user intent for the search query.
5. NLP in Content Editor
All the calculations take place on the backend. In your Content Editor query, you will see a separate tab on your Terms to Use list – both on the right-side panel and in the customization tab.
TIP: NLP Entities help understand Google your content better, so it's highly recommended to use those terms in your content!
6. NLP in Audit
In Audit, on your Terms to Use list, you will see an NLP tag next to some terms. If you enabled the sentiment upon creating your query, for those terms, we would also label the sentiment (negative/neutral/positive).