Intelligent term checks
NLP models for intelligent checks
To ensure that everyone in the company really enjoys working with Lexeri, it is essential that term checks are actually helpful and do not feel like additional work. That's why we incorporated trained language models from day one to show users the best possible results.
How good are the language models?
Lexeri builds on local language models that run on its servers and enable the use of various methods of Natural Language Processing (NLP).
For instance, Lexeri is able to recognise inflected forms of terms in textswithout them having to be added to your termbase
Among the methods we use for term extraction is Named Entity Recognition, which identifies particularly relevant terms in your texts. We also analyse the frequency with which the text refers to a term as a means of measuring its relevance.
Broadly speaking, NLP models deliver better results in term checks with fewer incorrect or unrecognised terms. This means in turn that your employees will take more pleasure in checking texts, use the option more frequently and hence improve the consistency of your documentation.