Arabic Readability Assessment

The BAREC Shared Task 2025, part of the Third Arabic Natural Language Processing Conference (ArabicNLP 2025) @ EMNLP 2025.

Strict Track

For the strict track, models must be trained exclusively on the BAREC Corpus.

Task 1: Sentence-level Readability

Predict readability level of an Arabic sentence on a scale from 1 to 19.

Task 2: Document-level Readability

Predict a document's readability level (1-19) based on its most difficult sentence.

Constrained Track

Participants may use the BAREC Corpus, SAMER Corpus, and the SAMER Lexicon.

Task 1: Sentence-level Readability

Predict readability level of an Arabic sentence on a scale from 1 to 19.

Task 2: Document-level Readability

Predict a document's readability level (1-19) based on its most difficult sentence.

Open Track

No restrictions on external resources, allowing the use of any publicly available data.

Task 1: Sentence-level Readability

Predict readability level of an Arabic sentence on a scale from 1 to 19.

Task 2: Document-level Readability

Predict a document's readability level (1-19) based on its most difficult sentence.