The first step is for search engines to understand what kind of information the user is looking for.
Analyze User Query
To do that, they analyze the user’s query (search terms) by breaking it down into a number of meaningful keywords.
A keyword is a word that has a specific meaning and purpose.
For example, when you type “How to make a chocolate cake”, search engines know from the words how-to that you are looking for instructions on how to make a chocolate cake and thus the returned results will contain cooking websites with recipes.
If you search for “Buy refurbished ….”, they know from the words buy and refurbished that you are looking to buy something and the returned results will include eCommerce websites and online shops.
Machine learning has helped them associate related keywords together. For example, they know that the meaning of the query “how to change a light bulb” is the same as this “how to replace a light bulb”.
They are also clever enough to interpret spelling mistakes, understand plurals and in general extract the meaning of a query from natural language (either written or verbal in case of Voice search).
