A Review of Advances in Computational Analysis of Kazakh Political Discourse

Abstract

This review article examines advances in computer analysis of Kazakh political discourse in the context of broader research in natural language processing, knowledge engineering, and digital humanities. It addresses three main issues: the digitisation of the public sphere, the morphological complexity of the Kazakh language, and the ongoing transition to the Latin alphabet. Rather than introducing new empirical data, the paper integrates and systematises existing studies into a reproducible methodological framework. The discussion highlighted achievements in several areas, including the establishment of an ethical and legal framework for obtaining text, the use of ontological modelling to structure campaign communication and enable semantic justification, the development of preprocessing strategies that take into account spelling variations and morphological ambiguity, as well as the introduction of subsequent methods such as sentiment analysis, stance. identification, semantic similarity, and framing analysis. An integrated methodological approach is proposed that combines ethical data collection, ontology development, language preprocessing, and computational interpretation into a single framework that emphasises reproducibility and comparative evaluation. Special attention is paid to the need to create annotated corpora, specialised dictionaries and reproducible workflows for the development of this developing field. By summarising methods, challenges, and opportunities, the article demonstrates how computer-based approaches can enrich the study of Kazakh political discourse and contribute to broader debates in the fields of political communication, computational linguistics, and digital humanities.

More by Altanbek Zulkhazhav

• Quantifying the Intelligibility–Expressivity Tradeoff in Emotion-Conditioned Kazakh Text-to-SpeechInternational Conference on Affective Computing and Intelligent Interaction Workshops and Demos · 2026
• Acoustic and Emotional Annotation of a Web-Harvested Kazakh Speech Corpus for Text-to-Speech2026
• Monitoring and comparative analysis of containerized environmentsFrontiers in Computer Science · 2026
• A Hybrid Attention-LSTM Model with Embedding-based Feature Representation for Student Academic Performance PredictionInternational journal of intelligent engineering and systems · 2026