Azure Search Now Supports Lucene Fuzzy Searching in Preview

Azure Search provides a cloud based text indexing engine – this blog uses Azure search for its search engine.  The underlying indexing engine is lucene.net wrapped in a cloud service.  The search query API is provided through a REST API and a .NET SDK.   

In the latest preview version of the Azure Search API, you can now execute Lucene queries that enable some new features for querying the search index:

  • Fuzzy Search: finds terms that are similar.  For example, “blue” is similar to “blue”, “blues”, and “glue”.
  • Proximity Search: finds terms that are in proximity but not right beside each other in a phrase.  For example, searching for “azure search” could return “azure search” but also “azure uses search” or “azure indexed by search” because azure and search are close to each other.
  • Term Boosting: allows you to boost one term over another in your results.  For example the search query “rock^2 music” will boost results containing the word rock higher than results containing the word music.
  • Regular Expression Search: allows for searching based on regular expression. 
  • Wildcard Search: allows for searching for terms that start with a specific query but then follow with multiple additional characters.  For example, “star*” would find documents containing “starlight”, “starship”, “starman”, etc.

In order to leverage these new APIs, you have to use the 2015-02-28 Preview API.  In addition, the .NET SDK doesn’t yet support this release, so you’ll have to construct your own REST calls.