In Part 3 of a series on data structures in NoSQL databases, the focus is on utilizing Full-Text Search (FTS) natural language queries with Couchbase. The post explains that FTS indexing in Couchbase records words from text-based document structures, allowing efficient retrieval of documents matching search terms. It contrasts FTS with N1QL queries, noting that FTS returns documents while N1QL returns rows. The text highlights Couchbase's use of inverted indexing for FTS and its ability to score and order results by relevance, updating indexes as new data is added. Couchbase's basic data structures, such as maps, lists, sets, and queues, are stored as JSON documents, with maps being exclusively indexable due to their field names. The post includes examples of creating and searching data structures using the Python SDK and describes Couchbase's capabilities for advanced querying, such as fuzzy matching and geospatial searches. It also touches on Couchbase's scope and collection-level searching introduced in version 7.0, emphasizing the platform's comprehensive nature and ease of use for developers building and managing data applications.