There are three components of time series that are key to understanding time-related data. They are trend, seasonality, and noise. Let's look at each of them in the context of our EU unemployment data.
The trend can be defined as the long-term tendency of the time series data—the fact that, on average, the values tend to increase or decrease over a period of time. Looking at our plot, we can identify three distinct trends:
A downward trend from 2005 until 2008 (less people unemployed on a year-on-year basis); an upward trend starting in 2008 and manifesting until 2013 (unemployment rose on average); and again, a downward trend between 2013, all the way until the end of 2017 (the number of people without work constantly decreased).