A Novel Coding and Discrimination (CODIS) Algorithm to Extract Features from Arabic Texts to Discriminate Arabic Poems

摘要

This article proposes a new algorithm to ‎discriminate Arabic poems by inserting Arabic poems ‎texts and coding Arabic letters, extracting letters features ‎depending on letter shapes to construct a multidimensional ‎contingency table, and analyses the frequencies of letters in ‎the inserted texts statistically. The proposed coding and ‎discrimination (CODIS) algorithm could be applied for ‎different Arabic texts in any media. A sample of five poems ‎for six poets was examined to implement a CODIS algorithm. ‎A Chi-Square statistic is used to determine the relation between ‎the features and discriminate poems.‎.