Think of students writing a book review for a class. If we have a book about social media, most people will probably include social media in their description. However, Peter uses the words "online social networks", and his classmate Mary discusses establishing a personal presence on the web. Computers look for exact matches when processing data. A computer searching through the class book reports for social media would not find either Peter's or Mary's book reviews because they did not use "social media" to describe the book.
Subject headings or controlled vocabulary are a way to ensure that everyone looking for information about social media finds all the material available. Database planners add a field to the item's description containing uniform, agreed-upon terms describing subject content. So the teacher in our example above would enter all his student book reviews into the computer adding the Subject heading: "social media" to each book review for the book. The teacher might also include book reviews that the same students had written on The Scarlet Letter and Harry Potter. They would have the subject headings "Literature" and "Fantasy." At the end of the semester, the teacher can use the Find function to find those reviews with the words "social media" in the document. Some reviews would contain social media in the student description (abstract), some would not, but all the relevant reviews would have the subject heading "social media" added by the teacher.
Note: Subject terms are consistent in the database you are in. They may vary database to database, so just be aware of this as you are using them in your research.