Suppose that you are a human search engine, trying to sort documents in an
appropriate order in response to a user query. Perhaps you would place the documents
in piles, like “good,” “not so good,” and “bad.”The computer is doing essentially
the same thing with scoring. However, you might also break ties by looking
carefully at each document to decide which one is more relevant. Unfortunately,
finding deep meaning in documents is difficult for computers to do, so search engines
focus on identifying good features and scoring based on those features.