EDUCATIONAL AND LINGUISTIC ASPECTS OF TEXT PREPROCESSING IN PASHTO

Abstract:This paper introduces foundational text preprocessing strategies designed for the Pashto language, focusing on linguistic structure, educational adaptation, and syntactic modeling. Given Pashto's complex morphology, diverse dialects, and lack of standardized digital tools, the study outlines a step-by-step pipeline including orthographic normalization, rule-based tokenization, customized stop-words filtering, and light stemming suited to Pashto syntax. The work further explores grammatical tagging and lexical refinement with emphasis on Pashto educational and literary terms. Rather than relying solely on machine learning models, this research emphasizes human-curated linguistic rules, making it suitable for academic, pedagogical, and resource-building contexts. The developed scripts and processed datasets serve as baseline tools for future Pashto NLP applications, corpus development, and digital resource integration. Results highlight the importance of tailored preprocessing in improving text quality, readability, and annotation consistency across domains.