Nati-Hab1/Text-Processing-on-Ethiopian-Language-Pairs

📊 Comparative Analysis of Amharic and Tigrigna - A Python-based project that measures linguistic similarity between Amharic and Tigrigna by analyzing word-level and phoneme overlap using SERA-based G2P tools. Results show high phonetic similarity and low lexical overlap, offering insights for linguists, educators, and NLP developers. # Comparing and Analyzing the Word-Level and Phoneme Overlap of Amharic and Tigrigna ## 📄 Abstract This project explores the linguistic similarities between Amharic and Tigrigna by analyzing word-level and phoneme overlap between the two languages. Utilizing a combination of Python programming and phoneme conversion tools, the study quantifies the degree of overlap, providing insights into the structural and phonetic relationships between these closely related Semitic languages. The results reveal significant phonetic similarities and notable word-level overlap, contributing to the understanding of the linguistic proximity between Amharic and Tigrigna. ## 🔍 Introduction Amharic and Tigrigna are both members of the Semitic branch of the Afro-Asiatic language family, spoken predominantly in Ethiopia and Eritrea. Despite their shared linguistic heritage, they exhibit distinct phonetic, lexical, and syntactic characteristics. This project aims to quantify the degree of similarity between Amharic and Tigrigna by comparing word-level and phoneme overlaps. By leveraging phoneme conversion algorithms and analyzing the resulting data, the study provides a clearer picture of the linguistic relationship between the two languages. --- ## 🧪 Methodology The methodology for comparing and analyzing the word-level and phoneme overlap involved several stages, all implemented in Python. ### 1. Data Preprocessing - Cleaned the text data by removing punctuation and non-phonetic characters using `clear_non_alphabet`. - Stored the cleaned text in a standardized format for further processing. ### 2. Phoneme Conversion - Applied a SERA-based grapheme-to-phoneme algorithm using `convert_to_phonemes`. - Stored phonetic representations for further phoneme-level analysis. ### 3. Word-Level Comparison - Used `frequency_counter` to analyze word frequency in both languages. - Computed: Similarity Percentage = (Number of Common Words) / (Total Unique Words - Common Words) × 100 - Saved re …