Design Retrieval Systems Using Advanced Chunking Strategies GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A team is comparing document-based chunking (by sections) and semantic chunking for a RAG system over a technical manual. They want to know which strategy yields better retrieval performance for user queries. They have a set of queries with known relevant sections. Which metric should they use to compare?

   • ARecall@k, measuring how many relevant chunks are retrieved in the top k results.
   • BThe average length of chunks produced by each strategy.
   • CThe time taken to embed all chunks.
   • DThe number of chunks that contain the word 'manual'.
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   How does a sliding window approach help in context-aware chunking?

   • AIt uses embeddings to determine the window size dynamically.
   • BIt ensures that each chunk has a fixed size by using a window of exactly the same length.
   • CIt splits the document into non-overlapping sections based on headings.
   • DIt creates overlapping chunks so that information at the boundaries is included in multiple chunks, preserving context.
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   How does semantic chunking determine where to split text?

   • ABy splitting at fixed character counts with overlap.
   • BBy measuring the embedding similarity between consecutive sentences and grouping similar ones together.
   • CBy splitting at every sentence boundary regardless of content.
   • DBy using the document's headings and subheadings as split points.
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   What does the 'recall' metric measure in the context of evaluating a retrieval system?

   • AThe proportion of all relevant chunks that are successfully retrieved.
   • BThe number of chunks stored in the vector database.
   • CThe proportion of retrieved chunks that are relevant.
   • DThe speed at which the retrieval system returns results.
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   Which of the following best describes the output of fixed-size chunking?

   • AChunks that are determined by the semantic similarity of sentences within the document.
   • BChunks that are created by splitting the document at every heading and subheading.
   • CChunks of varying lengths that align with natural language boundaries like paragraphs.
   • DChunks of a constant token or character length, with an optional overlap between consecutive chunks.
   Reveal Solution Discussion Select an answer first