Research Note
Search, Recommendation, and Generation System Map
Search begins with an expressed information need. The system interprets a query, retrieves candidates from an eligible corpus, and ranks them for that need.
In this article
Search, Recommendation, and Generation System Map
Search begins with an expressed information need. The system interprets a query, retrieves candidates from an eligible corpus, and ranks them for that need.
Recommendation predicts which items may be useful or engaging from user, item, context, and interaction signals. It may have no explicit query. Production recommenders commonly separate candidate retrieval from ranking and post-ranking.
Generation can explain, summarize, or synthesize over retrieved material, but it does not replace corpus eligibility, retrieval, ranking, provenance, or evaluation. Similar interfaces can hide different discovery systems, so product language should name the actual path.
Sources
Follow the evidence.
- Introducing Llama 3.1ai.meta.com
- ai.meta.com: the llama 3 herd of modelsai.meta.com
- NIST AI Resource Centerairc.nist.gov
- csrc.nist.gov: finalcsrc.nist.gov
- daltonanderson.ghost.io: metas ai power play llama 3 smart reel searchdaltonanderson.ghost.io
- Meta Llama models repositorygithub.com
- github.com: LICENSEgithub.com
- github.com: MODEL CARDgithub.com
- github.com: USE POLICYgithub.com
- open.spotify.com: 5xmE0hYheRvBOoqaQCyUokopen.spotify.com
- elastic.co: search rank evalelastic.co
- etsi.org: 2457 etsi releases new guidelines to enhance cyber security for consumer iot devicesetsi.org
- nist.gov: 7 tips keep your smart home safer and more private nist cybersecuritynist.gov
- NIST AI Risk Management Frameworknist.gov
- tensorflow.org: Retrievaltensorflow.org
- tensorflow.org: basic retrievaltensorflow.org
- tensorflow.org: recommendation systemstensorflow.org
- youtu.be: J2I1fJW1sB4youtu.be