Search Relevance is the practice of manipulating a search application's behavior to improve the quality of the search results based on the needs of the application's users
Search Relevance is the practice of improving the quality of search results in a search application. In relevance work, you line up a user audiences needs (as expressed in their natural language search string) with the content available in the search application.
For example, a customer typing the search "cancer" into a hospital site's search is likely to have very different expectations than doctors typing that term into a medical research site. In the hospital case, users want a "info desk" search user experience, and are likely interested in the hospital's cancer services. For doctors searching a research site, they're more likely to be interested in the latest cancer research.
With search relevance, you work to understand the user's needs and line up the search application's ranking behavior to match. Relevance work involves technical work to manipulate the ranking behavior of a commercial or open source search engine like Solr, Elasticsearch, Endeca, Algolia, etc. This means manipulating field weightings, query formulations, text analysis, and more complex search engine capabilities. It may also mean leveraging user behavioral data, NLP, statistical, and other machine learning techniques to modify or enrich the behavior of such a search engine.
More Reading
Blog
Books
- Major part of the larger topic of Information Retrieval of which Intro to Information Retrieval is the canonical tome
- Relevant Search by Manning Publications -- more practical Solr/Elasticsearch relevance book
