Unstructured Data

Unstructured data is information without a predefined format or schema. Examples include free-text product descriptions, customer reviews, property listing copy, images, PDFs, and audio recordings. Unlike structured data, you cannot directly query a specific field - the information must be extracted or processed first.

APIs that include user-generated content - product reviews or property descriptions - return unstructured text fields alongside structured metadata. Analyzing this content for sentiment, keyword extraction, or summarization typically requires natural language processing (NLP) or AI models.

The boundary between structured and unstructured data is not always sharp. A review dataset might have a structured rating field alongside an unstructured reviewText field, requiring different processing approaches for each.