You are here

Connexor Search Technologies

The vast amount of information around us have made conducting a search a rather challenging task. Connexor's technologies help to overcome some of these challenges in text search:

Inflection

In the traditional approach to information retrieval, the search system looks for the exact matches of the search query. However, in many languages the words may undergo a change to mark such distinctions as case, number, tense, or person. As the keyword or the phrase may appear in the text in inflected form, a simple search cannot always find all the matches. Connexor solves the problem by using baseforms in queries. This allows for example to match a Finnish word laki (= law) and its inflected forms like lakien (= laws') and lailla (= by law).

Compounds

When search query is a compound word, we might also be interested in finding matches to a part of a compound word. For example, if searching with the German language word Eisenbahn (=railway), one may like also to find those occurrences of the word Bahn (=track). Connexor technology finds the boundaries between the compound parts and thus allows you to make searches also with these compound parts.

Synonymy and abbreviation

Concepts are often identified by multiple terms. Names are sometimes abbreviated to save space or effort. In text search, it is thus desirable that the search results include all the other words with exactly the same meaning. In order to find the relevant information more accurately, Connexor search technology allows to associate synonyms and to find matches between abbreviations and corresponding unabbreviated words or phrases.

Finding the relevant content of a document

For accurate search results, it is crucial to know what the most relevant content of the document is. The sophisticated metadata discovery technology offered by Connexor is one, effective way to ensure that the search system retrieves the most relevant documents matching the search criteria. Language identification and document classification narrow down the group of valid documents. When displaying search results automatic summarization allows a quick generation of brief overlook on the document content.