This study extracts eleven types of attributes from English text data in order to classify English text by learning and categorization level of difficulty. Text is submitted to machine learning and categorization using the “leave-one-out cross-validation” method. E-books have recently become more popular. The process of manually categorising all of the e-books takes a long time as the number of them grows. It is possible to recommend a foreign-language book that is appropriate for the reader’s level of English ability when English sentences are categorised according to their difficulty level. Furthermore, an experiment is conducted in which the size of text data is adjusted and the attribute selection approach is performed in order to increase accuracy. As a result, accuracy has increased to 77.04 percent, and the F-measure has increased to 63.96 percent. Erroneous identification is also noticed as a result of the influence of columns between sentences.
Faculty of Engineering, Sanjo City University, Niigata, Japan.
Graduate School of Natural Science and Technology, Kanazawa University, Ishikawa, Japan.
Faculty of Production Systems Engineering and Sciences, Komatsu University, Ishikawa, Japan.
Please see the link here: https://stm.bookpi.org/CRLLE-V4/article/view/6157