Work

SUBTLEX-AR: Arabic word distributional characteristics based on movie subtitles

Year: 2025

Type: article

Abstract: This article presents SUBTLEX-AR, a digital database providing an extensive collection of attributes related to Modern Standard Arabic words (Arabic for short). SUBTLEX-AR combines a novel dataset of 120 million word tokens from movie subtitles with 40 million tokens from newspaper articles originally collected in ARALEX (Boudelaa & Marslen-Wilson, Behavior Research Methods, 42, 481-487, 2010), ensuring comprehensive coverage. SUBTLEX-AR provides information about the statistical properties of Arabic words at the orthographic, phonological, morphological, and semantic levels. The database also includes information on sub-word structure properties like bigram and trigram frequencies, as well as lemmas and part-of-speech information along with their corresponding frequencies. The online interface of SUBTLEX-AR allows users either to upload a set of words to receive their properties or to receive a set of words matching constraints on predefined properties. The properties themselves are easily extensible and will be expanded over time. SUBTLEX-AR is freely accessible here: https://subtlexar.uaeu.ac.ae/.

Source: Behavior Research Methods

DOI: 10.3758/s13428-024-02560-8

Authors: Sami Boudelaa, Manuel Carreiras, Nazrin Jariya, Manuel Perea