Hybrid Analysis for Karakalpak Language: Combining Statistical Model and Rules-Based Approach

This paper presents a hybrid approach to morphological and syntactic analysis of Karakalpak texts. The proposed method combines traditional linguistic rules and statistical models based on n-grams, which allows for effective resolution of ambiguities at both morphological and syntactic levels. Morphological analysis is implemented using dictionaries and an iterative affix removal mechanism covering 144 unique affixes of the Karakalpak language. For syntactic analysis, frequency lists of bigrams and trigrams compiled on the basis of a corpus of $\mathbf{1 0, 0 0 0}$ sentences are used to select the most probable syntactic structure of sentences. The results of the study confirm the feasibility of the proposed approach for resourceconstrained languages and show the potential for its use in applied natural language processing. Besides, authors conducted quite deep research about comparative analysis of existing solutions. Moreover, there is also useful and important information about Karakalpak language and its nature, which helps to understand its limitations and difficulties during development of processing algorithm (tool).