Linguistics for Elasticsearch & OpenSearch

For your customers and users, searching almost always poses a big challenge: they never formulate their search queries so perfectly that Elasticsearch can understand and process them without errors. This is because of singular and plural word forms and compound phrases, which are simply par for the course for somewhere more complex languages. Elasticsearch does not have the “linguistic understanding” for this, and the software is therefore only rarely able to produce the perfect search result with suitable hit lists and suggestions. Too bad for you because the risk to frustrate or scare off your customers/users is proven to be very high.

But now you have the chance to easily serve your users better, because we have the perfect solution for your Elasticsearch installation: for almost two decades we have been working on linguistic and morphological products that understand language better and deliver better search results to hundreds of thousands of users daily. We’ve packed the concentrated power of this experience into this linguistic improvement, a “tweak” for Elasticsearch. It allows you to improve your Elasticsearch installation by simply integrating a plug-in.

Even better: your users will find what they're looking for with greater ease And you convince with your application!

By installing the linguistic plug-in, you are harnessing the power of technologies that noticeably improve the recall (completeness) and precision (relevance) of your search engine: your Elasticsearch platform gets an upgrade that helps your users/customers find what they’re looking for and never overlook something relevant again.

The following technologies are integrated in this solution: starting from a professional identification of language and word boundaries – also of complex terms – a very precise morphological base form normalization process measurably improves search quality. Thereby the search for "author" also delivers hits with the word "children's book author". A search for "book" will also provide hits for declined forms such as "books".

The decomposition of arbitrarily composed terms into meaningful single components enhances the quality as well: The search for "book" also delivers hits with "children's books", the search for "Federal Research Ministry" delivers hits with "Federal Ministry for Science and Technology". Due to the infinite number of possible combinations in the formation of compounds, the IntraFind linguistics tool works with intelligent filter rules and procedural lexicons.