Claim detection is the first step in the fact-checking pipeline. It focuses on identifying whether a given sentence or piece of text expresses a verifiable factual claim or not. This process distinguishes between statements that can be objectively verified and those that cannot (e.g. opinions).
The task involves classifying input text into two main categories:

• Claim: The sentence contains a verifiable factual claim.
• Non-Claim The sentence does not express a verifiable claim and may include personal opinions, questions, or vague/general statements.

This step plays a critical role in reducing the workload for fact-checkers by automatically identifying sentences that are potentially check-worthy. By focusing only on factual claims, fact-checkers can prioritize their efforts and avoid wasting time on unverifiable content.