4-level science readability models struggle most with distinguishing adjacent grades 1 and 2¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2024 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Confusion-matrix analysis of the 4-level models shows that, for all three languages, both ML and DL models most often confuse level 1 and level 2 documents; the Basque model also struggles with levels 1 and 3. → Uçar 2024

Evidence¶

Uçar 2024¶

Uçar, S.-Ş., Aldabe, I., Aranberri, N., & Arruarte, A. (2024). Exploring Automatic Readability Assessment for Science Documents within a Multilingual Educational Context. Int J Artif Intell Educ. https://doi.org/10.1007/s40593-024-00393-2

q2 · i? · design · r2

Closer inspection of confusion matrices for the 4-level cross-validation models across Basque, Spanish and English. 'No considerable differences or patterns can be observed among the remaining level-pairs for any of the models.'

"We observed that, for the three languages in the 4-level setup, both the feature-based ML and the DL models struggle the most when deciding whether a document belongs to level 1 or level 2."

Discussion¶

Related Claims¶

• Science readability models classify texts substantially more accurately in a 2-level setup than in the 4-level ESO-grade setup — related
• Linguistic complexity features show few clear progressions across ESO grades in the compiled science corpora — related