Stop Word in Readability Assessment of Thai Text

Teachers and parents may use readability to select appropriate learning materials for primary school students. This research constructs Thai stop word list and evaluates the impact of eliminating stop words on readability assessment of Thai text. The corpus contains 1,188 textbook articles used by students from grade 1 to grade 6. Word segmentation, stop word list extraction, and feature selection are the preprocessing tasks performed on the articles in the corpus. Then, term frequency and inverse document frequency (TF-IDF) of the selected terms are used as features for support vector machines (SVMs) to generate classification models. Experimental results show that F-measure can reach 0.87 when identifying Thai articles suitable for middle grades primary school students.