A software company receives thousands of free-text customer support tickets each day. The support manager wants to automatically identify the most important terms and topics in each ticket so agents can quickly understand what a ticket is about without reading the full text. Which text-analysis technique should be applied?