Adding this context to raw data is a process called data labeling, considered a key step in training machine learning algorithms.