دانلود کتاب مورفولوژی محاسباتی حالت محدود

Georgian presents a lot of different morphological features and is a highly interesting language for NLP systems, especially from the viewpoint of computational morphology. Modeling Georgian presents us with an opportunity to adapt and to test computational techniques, to build applications and to develop interdisciplinary approaches to languages with rich morphology. This book can be seen as an attempt to demonstrate how technology can be used for the processing of highly agglutinative language types like Georgian. This book also brings together most of the work done in Georgian in the last 8 years. It covers the compilation of corpora of Georgian texts and the development of tools including a tokenizer, a rule-based morphological analyzer, and a generator of Georgian. The computational model presented here, which is implemented by means of finite-state technology, specifically lexc and xfst, encodes the morphosyntactic features of Georgian language, while the transducer, which has been tested on the Georgian Language Corpus, can be used as a component of other NLP applications for lemmatizing or tagging other resources. It is also our hope that this book will assist other researchers in advancing their knowledge of Georgian and, possibly of other Kartvelian languages sharing similar features, and in using technology for their processing activities.