Since October last year, Textkernel released deep learning in English, French and German to its CV parsing algorithm, Extract! 4.0. This Month’s release contains also Dutch, Flemish, Spanish, Russian and simplified Chinese parsing.
Automated CV processing
For us people, reading a CV is fairly simple. CV’s generally contain clearly recognizable and predictable parts for people. If a CV has a different layout than we are used to, or if it contains other information than we expect, we are still able to extract all relevant information from the document. Computer programs, on the other hand, have to be continuously developed to ensure that data is extracted from CV’s with the highest possible accuracy.
Automated ‘reading’ of data from CVs is also called CV parsing. The technology recognizes the (unstructured) data in a CV, structures it and stores it searchable in any desired software system. Manual data entry – by recruiter or candidate – is no longer necessary, it improves the candidate experience and forms the basis for an optimal semantic search and matching process.
With a fully based on deep learning CV parser, Textkernel is one of the leading research and development laboratories in the field of artificial intelligence (AI) for recruitment and recruitment and selection, a new era for recruitment. The implementation of Deep Learning technology in the Dutch CV model makes the parser even more accurate: extraction errors have decreased by as much as 15% to 30%. Already in the English Extract! 4.0 were greatly improved results.
What is deep learning
Deep Learning are AI algorithms that automatically find patterns in data using brain-like structures called Neural Networks. From self-driving cars to automated personal assistants, Deep Learning has revolutionized nearly every fields it has been applied to. It has in the past few years transformed entire industries including automotive, healthcare, retail and financial services.