Computer based structural analysis of Turkish words

A method and a software package, based on this method, for the structural analysis of Turkish words are explained. The steps for the structural analysis of a word are as follows: determine the root of the word using string matching instructions and using the related root bank; partition the suffix group into suffixes using the related data files and the grammatical rules; extract the proper meaning explanation for the suffixes; and table out the results. The average analysis time for a given word with two suffixes, which is the most common usage in Turkish, is 0.05 seconds for an IBM-AT compatible computer which runs at 50 MHz. For 9000 roots the developed software, together with the data files, occupies 310 KBytes. Compared with the 60 MBytes of the storage area which would be used in the case of direct storage of the words derived from these roots, the method suggested and used by this study is a great improvement from the point of view of storage.<>