Definition
Latent Semantic Indexing (LSI) uses word associations (words related to the subject of a search as opposed to the actual keywords entered) to help search engines know more accurately what a webpage is about. [1][2]
from datacamp.com
Also known as latent semantic analysis, LSI is a mathematical practice that helps classify and retrieve information on particular key terms and concepts using singular value decomposition (SVD), which enables search engines to scan through unstructured data and identify any relationships between these terms and their context to better index these records for users online. Before SVD, it was rather difficult for computers to try and grasp differences between synonyms or semantic changes. [3]
References
- American Marketing Association. AMA Dictionary.
- SEMPO (Search Engine Marketing Professional Organization). SEM Glossary.
- Hubspot, What is Latent Semantic Indexing & Why Does it Matter for Your SEO Strategy?