Evaluating Various Tokenizers for Arabic Text Classification

The first step in any NLP pipeline is to split the text into individual tokens. The most obvious and straightforward approach is to use words as tokens. However, given a large text corpus, representing all the words is not efficient in terms of vocabulary size. In the literature, many tokenization algorithms have emerged to tackle this problem by creating subwords, which in turn limits the vocabulary size in a given text corpus. Most tokenization techniques are language-agnostic, i.e., they do not incorporate the linguistic features of a given language. Not to mention the difficulty of evaluating such techniques in practice. In this paper, we introduce three new tokenization algorithms for Arabic and compare them to other three popular tokenizers using unsupervised evaluations. In addition, we compare all the six tokenizers by evaluating them on three supervised classification tasks: sentiment analysis, news classification and poem-meter classification, using six publicly available datasets. Our experiments show that none of the tokenization techniques is the best choice overall and that the performance of a given tokenization algorithm depends on many factors including the size of the dataset, nature of the task, and the morphology richness of the dataset. However, some tokenization techniques are better overall as compared to others on various text classification tasks.