ElasticSearch Research & Evidence-Based Practice 4 — Questions and Answers
Question 1: A research team needs to enrich ingested papers by extracting named entities (author names, institutions) before indexing. Which Elasticsearch component handles data transformation at ingest time?
- Search template
- Ingest pipeline with processors (Correct answer)
- Index lifecycle policy
- Watcher alert
Correct answer: Ingest pipeline with processors
Ingest pipelines with processors (e.g., grok, set, script) transform documents before they are indexed, enabling field extraction and enrichment at write time.
Question 2: An evidence synthesis platform needs to group search results by author and show only the most recent paper per author. Which Elasticsearch feature supports this?
- terms aggregation with top_hits sub-aggregation
- collapse field with inner_hits (Correct answer)
- cardinality aggregation
- sampler aggregation
Correct answer: collapse field with inner_hits
Field collapsing deduplicates search results by a specified field and uses inner_hits to expose additional documents within each collapsed group.
Question 3: A researcher discovers that some indexed papers have incorrect 'study_type' values due to a classification error. They need to update this field across thousands of documents matching a filter. Which API is most appropriate?
- Bulk API with index action
- Update by Query API (Correct answer)
- Reindex API
- Update API with _id
Correct answer: Update by Query API
Update by Query applies a script or partial update to all documents matching a query, making it the right tool for bulk field corrections without reindexing.
Question 4: A systematic review tool needs to rank papers by a combination of citation count (numeric) and BM25 text relevance. Which Elasticsearch feature supports combining a numeric signal with text score?
- rescore query
- function_score query (Correct answer)
- dis_max query
- boosting query
Correct answer: function_score query
function_score query wraps another query and applies custom scoring functions (e.g., field_value_factor) to blend numeric signals with the base relevance score.
Question 5: Researchers need a persistent, named search that can be executed repeatedly against evolving data without resubmitting the full query. Which Elasticsearch feature supports this use case?
- Saved search in Kibana only
- Stored search template (Correct answer)
- Async search API
- Search pipeline
Correct answer: Stored search template
Stored search templates allow parameterized queries to be saved in Elasticsearch and reused via the _search/template endpoint with different parameters each time.
Question 6: A research dataset index grows daily with new publications. To keep only the last 12 months of data online and move older data to cheaper storage, which Elasticsearch feature automates this lifecycle?
- Snapshot lifecycle management
- Index lifecycle management (ILM) with rollover and delete phases (Correct answer)
- Curator scheduled deletion job
- Watcher time-based trigger
Correct answer: Index lifecycle management (ILM) with rollover and delete phases
ILM automates index transitions through hot, warm, cold, and delete phases based on age or size, making it the standard way to manage data retention policies.
Question 7: A clinical evidence database needs to ensure that only administrators can delete or modify indexed studies, while analysts can only read. Which Elasticsearch security feature enforces this?
- Document-level field masking
- Index-level privileges in role definitions (Correct answer)
- Ingest pipeline access control
- Search template permissions
Correct answer: Index-level privileges in role definitions
Elasticsearch role-based access control uses index-level privileges (read, write, delete, manage) assigned per role to control who can perform which operations on an index.
ElasticSearch Research & Evidence-Based Practice 4
A research team needs to enrich ingested papers by extracting named entities (author names, institutions) before indexing.
Which Elasticsearch component handles data transformation at ingest time?