Many times a year, Google updates the algorithms in a way that usually describes an improvement to how Google understands search queries and content of a website. These changes affect how website content is seen and measured by search engines, which in turn affects the numbers of visitors to your page.
A number of in-depth research papers have been written about how Google interprets individual words, context, and long-form content to predict what someone is looking for. These papers assess the Google RankBrain and Neural Matching systems among others.
These papers combined are an excellent source of information for building web content and are well worth reviewing.
The ultimate takeaway from the research is that synonyms are important in content but still need to be used naturally and that Google assesses content at three different levels – word, sentence and paragraph. Content still needs to communicate an overall core meaning. Content quality is still an important ranking factor for Google.