ReadMe++: Benchmarking Multilingual Language Models for Multi-Domain Readability Assessment
This addresses the problem of limited evaluation resources for multilingual and multi-domain readability assessment, enabling more robust research in this area.
The paper tackles the lack of domain and language diversity in readability assessment by introducing ReadMe++, a dataset of 9757 sentences in five languages from 112 sources, and benchmarks models showing superior domain generalization and cross-lingual transfer.
We present a comprehensive evaluation of large language models for multilingual readability assessment. Existing evaluation resources lack domain and language diversity, limiting the ability for cross-domain and cross-lingual analyses. This paper introduces ReadMe++, a multilingual multi-domain dataset with human annotations of 9757 sentences in Arabic, English, French, Hindi, and Russian, collected from 112 different data sources. This benchmark will encourage research on developing robust multilingual readability assessment methods. Using ReadMe++, we benchmark multilingual and monolingual language models in the supervised, unsupervised, and few-shot prompting settings. The domain and language diversity in ReadMe++ enable us to test more effective few-shot prompting, and identify shortcomings in state-of-the-art unsupervised methods. Our experiments also reveal exciting results of superior domain generalization and enhanced cross-lingual transfer capabilities by models trained on ReadMe++. We will make our data publicly available and release a python package tool for multilingual sentence readability prediction using our trained models at: https://github.com/tareknaous/readme