Search

Text readability is one of the main research areas widely developed in several languages, but it is highly limited when dealing with the Arabic language. The main challenge in this area is to identify an optimal set of features that represent texts and allow us to evaluate their readability level. To address this challenge, we propose in this study various feature selection methods that can significantly retrieve the set of discriminating features representing Arabic texts. The second aim of this paper is to evaluate different sentence-embedding approaches (ArabicBert, AraBert and XLM-R) and compare their performances to those obtained using the selected linguistic features. We performed experiments with both SVM and Random Forest classifiers on two different corpora dedicated to learning Arabic as a foreign language (L2). The obtained results show that reducing the number of features improves the performance of the readability-prediction models by more than 25% and 16% for the two adopted corpora, respectively. In addition, the fine-tuned Arabic-BERT model performs better than the other sentence-embedding methods, but it provided less improvement than the feature-based models. Combining these methods with the most discriminating features produced the best performance. [ABSTRACT FROM AUTHOR]

The focus of data scientists is essentially divided into three areas: collecting data, analyzing data, and inferring information from data. Each one of these tasks requires special personnel, takes time, and costs money. Yet, the next and the fastidious step is how to turn data into products. Therefore, this field grabs the attention of many research groups in academia as well as industry. In the last decades, data-driven approaches came into existence and gained more popularity because they require much less human effort. Natural Language Processing (NLP) is strongly among the fields influenced by data. The growth of data is behind the performance improvement of most NLP applications such as machine translation and automatic speech recognition. Consequently, many NLP applications are frequently moving from rule-based systems and knowledge-based methods to data-driven approaches. However, collected data that are based on undefined design criteria or on technically unsuitable forms will be useless. Also, they will be neglected if the size is not enough to perform the required analysis and to infer the accurate information. The chief purpose of this overview is to shed some lights on the vital role of data in various fields and give a better understanding of data in light of NLP. Expressly, it describes what happen to data during its life-cycle: building, processing, analyzing, and exploring phases. [ABSTRACT FROM AUTHOR]