Product
Multilingual Search
Coreon Search is based on ESTeam ALiSe which runs on top of Lucene / Apache Solr – the proven enterprise search engine. ALiSe boosts Apache Solr’s matching capabilities with linguistic intelligence.
Wildcard and regular expression based techniques fail in morphological variation
treat ies > treat yor word order changes.
communication, telephonic > telephonic communicationFuzzy matching techniques based on permuting letter triples work only well with smaller amounts of terms; in large terminology collections too many hits are retrieved.
Coreon overcomes the boundaries of string / pattern-matching based search and goes way beyond brute-force fuzzy matches.
Linguistic intelligence
Multilingual Search
Linguistic intelligence
Morphology
Through compound analysers and built-in word nets you will find concepts even when the terms are missing from your resource – be surprised when searching for, say, dried flowers finds dehydrated plants.
Text Annotation
Term Recognition
Spot terms in sentences. Instead of manually typing in a keyword to look up a presumably existing term, simply send the complete sentence as a query: Coreon Search then scans the whole sentence, compares it to your data and highlights all found terms. Apply that when working in CAT or text annotation tools to guarantee consistency of your texts.
Yes, even with large graphs covering hundreds of thousands of terms. Coreon’s term search runs on Tantivy, a modern full-text search engine built for speed. Results appear almost instantly, and you can search in any language and find matching concepts across all languages in your graph.
Term recognition is just as fast. Instead of looking up terms one by one, you can send a complete sentence or text passage. Coreon’s recognition engine scans it in a single pass and highlights every known term. It stays quick even as your resource grows, which makes it ideal for real-time checks in translation, authoring environments, text annotation and AI/LLM-based chatbots.
Yes. You can narrow your search by:
- Language: Only look up results in the languages you work with.
- Branch: Search within one part of the knowledge graph, for example a single product line.
- Property: Filter by any of your properties, such as usage status.
- Date: Find entries created or changed in a given period.
Filters can be combined and shared with other users, so one can quickly pull up only approved terms in their language, and a team can work from the same filtered view.