CEFR Labelling and Assessment Services

Abstract Our pilot project aims to develop a set of text collections and annotation tools facilitate the creation datasets (corpora) for development AI classification models. These models can automatically assess text’s reading difficulty on levels described by Common European Framework Reference (CEFR). The ability accurately consistently readability level texts is crucial authors (language) teachers. It allows them more easily create discover content that meets needs students with different backgrounds skill levels. Also, in public sector using plain language written communication becoming increasingly important ensure citizens access comprehend government information. EDIA already provides automated assessment services (available as APIs an online authoring tool) CEFR English. Support Dutch, German Spanish are added part this project. Using infrastructure developed effort creating high quality additional languages lowered significantly. deployed through Language Grid. scheduled be completed second quarter 2022.