Google's ranking algorithm is a black box, but what we know for sure is that creating content that is topically relevant to the searchers search query is by far the most important factor.
As a writer, you should focus on making sure that your content is as relevant as possible to what your searchers are looking for.
But how do you know what's relevant to include and what's not?
That's where Frase comes in.
Frase recommends topics by analyzing the top Google results for your search query, and performing NLP analysis over them.
In a nutshell, Frase extracts content from each URL, and then leverages proprietary Named Entity Recognition to extract topics.
More specifically, this is how Frase goes from search query to list of topics:
input search query
process top 20 Google Search results and extract key topics per page
aggregate topics as one list across all results, and collect topic frequencies
automatically group topics into semantic clusters, while removing low quality outliers
various ranking factors are considered, including term frequency, cluster relevance, among others
score all pages with an algorithm that measures topic coverage