Then, individual morphemes can be further analysed to identify the grammatical structure of the word. In this paper, we propose a deep learning approach for learning the rules for identifying the morphemes automatically and segmenting them from the original word. Therefore, determining the morpheme boundaries becomes a tough task, especially in languages like Malayalam. Due to sandhi, many morphological changes occur at the conjoining position of morphemes. Sandhi splitting is important in the morphological analysis of agglutinative languages like Malayalam, because of the richness in morphology, inflections and sandhi. This process is termed as sandhi splitting. In order to perform this task, morphemes have to be separated from the original word. It is the study of the rules of word construction by analysing the syntactic properties and morphological information. Morphological analysis is one of the fundamental tasks in computational processing of natural languages.
0 Comments
Leave a Reply. |