Search engines put a lot of effort into identifying a page’s noun phrases and scoring each one in relation to the others found within the same textual piece of content. Noun phrases are at the top of a linguist’s arsenal. Studying the noun phrases and noun compounds in an article is an important part of computing fields called information retrieval and entity extraction.
Most keywords content editors and marketers try to target fall into one of two categories:
- Nominal phrases — a noun as head token completed with its modifiers and other nouns
- Verb phrases — a syntactic unit comprised of a verb and its dependent words
Nominal phrases, or noun phrases, are the most common syntactic unit in most textual content. The root is made up of a noun, and it can be modified by other nouns, adjectives, qualifiers, and determiners. While a single noun is
dog, a noun phrase could be
a fast hunting dog.
Verb phrases, on the other hand, have a verb as root token and all of its dependencies. For example,
hunts is a verb and a verb phrase could be
the dog hunts the small fowl. It is often a subject-predicate-object situation where we have a subject, the action performed, and what it is performed on. You can see on this verb phrase example that
the small fowl is a nominal phrase, too.
A noun phrase is a group of words comprised of a nominal head token, and acting as a subject, object, or prepositional object in a sentence. Also referred to as a nominal phrase, such sequence of words is often used in computational linguistics in order to extract the main topics discussed in a textual piece of content (e.g. transcribed video, blog post, research paper, etc).
The structure of a noun phrase can be broken down into different parts depending on what application you are interested in:
- article — e.g. the, a, our…
- quantifiers — e.g. some, a bunch of, many, ten…
- modifiers — e.g. best, luxury, dog, aluminum, heavy-duty…
- head token — e.g. dog, dog clippers, laptop…
- prepositional phrase — e.g. in London, for professionals…
Noun phrases and their saliency within a body of text have been studied over the years, if not decades. Indeed, because they give away so much information on what a document is about, many techniques have been elaborated to know which ones matter, and which ones should be dismissed. Once you have the right list of candidates noun phrases, you need to clean each one up. This class of tasks entails removing words that bring no further information about the subject matter, including:
- stopwords — e.g. incredible car dryers
- articles — e.g. the best car dryers
- arbitrary parts — not every marketer will care about the same parts of a noun phrase
The entire list of noun phrases from a blog post, or any document for that matter, conveys most of the meaning from that given text. While verbs and other lexical items matter very much, paying attention to your head tokens and the adjectives surrounding them will matter hugely for search engines at indexing.
In a sentence, noun phrases often represent either the subject or the object. In English, and this is also true in other languages, objects and subjects are often the most important parts of the contents. When you remove pronouns and determiners to then list all noun phrases, that somewhat summarizes the essence of the text at hand.
Verbs are great for intent and actions but the targets of these actions are defined using noun phrases. All modern search engines work hard on their part-of-speech tagging algorithms in order to identify, as precisely as possible, nouns and modifiers in a text. That way, they can offer very precise search results, and then focus on verbal phrases to catch the intent of the document being indexed.
As a content editor or content writer, your initial task is to shortlist the main head tokens for your article’s subject matter. That list could include only one noun, or seven, or more; this really depends on how long your article is going to be. Once this list of root nouns is established, choose the right modifiers for each. These modifiers will broaden up, narrow down, or bound the scope of your head token(s). Generally, they are either adjectives or nouns.
For example, a head token could be
laptop and its selected modifiers could be
aluminum. During the writing, you will want to put the emphasis on the head token and surround it with its modifiers. Avoid always using the same modifier unless your article requires it. If you write a blog post about
professional-grade laptops, you will inevitably use this noun phrase repeatedly; and that’s totally fine.
If you are a customer of the topicseed platform, you will see an advanced syntactic interpretation of every document you have in your library. Our machine learning-powered interpreter gives each word (token) its part-of-speech tag and analyzes each noun phrase, verbal phrase, and their dependencies. That way, you can quickly see what words get the most attention in your article, and improve your writing style to become more pertinent, and to-the-point.