If we want to represent a continuous signal (such as sound) in a computer, one way to do it is sampling.
This is the process of measuring the value (air pressure for sound) at specific times, often at a regular interval.
This interval is called the sampling frequency or sample rate.
In practice, if the sample rate is high enough, we can perfectly reconstruct the original signal.
You can see this intuitively on the chart on the right; you can still easily see the shape of the original signal even if you toggle its visibility.
Note: the example signals here are all simple sine waves, but this applies to more complex signals as well.