Predictive search is a concept for online search that aims to predict a user's search needs. It makes suggestions about search queries as soon as they are entered into a search bar.
The goal is to save the user time by allowing them to click on the suggestions instead of completing their search query themselves.
The core function of predictive search consists of autocomplete, autosuggest, or find-as-you-type.
For example, Google takes into account and gives preference to search queries from the user's previous search history and also popular search queries.
There are other approaches to successful predictive search, such as semantic search, which attempts to determine the meaning or intent behind the search query using contextual information.
Instead of focusing on the specific keyword, a semantic search engine with a predictive approach would suggest not only results that contain the keyword, but also those that contain synonyms.