in reality things are more complicated. You'll probably have thousands of books and customers. Each book now comes, not with a pair of numbers, but with a huge long list containing the rating of each customer or perhaps a blank if a specific customer hasn't rated the book. Now you can't simply plot the data and spot the pattern.
This is where topology comes to the rescue: it gives a neat way of turning shapes into networks. Suppose you've got a wobbly circle as in the figure below. You can cover it by overlapping regions and then draw a dot on a piece of paper for each region. You then connect dots corresponding to overlapping regions by an edge. The network doesn't retain the wobbliness of the shape, that information has been lost, but its topology, the fact that it's circular, is clearly visible. And the great thing is that it doesn't matter what kind of covering you use to make your network. As long as the regions are small enough — the resolution is high enough — the network will draw out the topology of the shape.