Amsterdam, 9 May – Textkernel has launched Extract! 4.0 – powered by Deep Learning – for German CV parsing. The new release, a major breakthrough in accuracy after more than a decade of R&D in the fields of Artificial Intelligence (AI), Machine Learning and Natural Language Processing, follows the initial release of Extract! 4.0 for English CV parsing in late 2017.
German CV parsing customers now benefit from even more accurate data, with better search results through improved extraction. Deep Learning technology has already achieved outstanding results in our English CV parsing model, and we’re pleased to announce that German CV parsing achieves improvements that are just as impressive. The German Extract! 4.0 model experiences increased accuracy across a large range of fields with error rates decreasing by 15 to 25 percent on average.
The Deep Learning enhancements of Extract! 4.0 were first launched in late 2017 for English CV parsing. German is the second language to benefit from these improvements, with more language upgrades planned in the future.