Normalisation of Indonesian-English Code-Mixed Text and its Effect on Emotion Classification

Usage of code-mixed text has increased in re-cent years among Indonesian internet users, who often mix Indonesian-language with English-language text. Normalisation this into needs to be performed capture the meaning English parts and process them effectively. We improve a state-of-the-art Indonesian-English normalisation system by modifying its pipeline modules. further analyse effect on emotion classification tasks. Our approach significantly improved normal-isation both individual modules overall system. The new feature set language identification module showed an improvement 4.26% terms F1 score. combination machine translation ruleset lexical BLEU score 25.22% lowered WER 62.49%. use context 2.5% 8.84%. effectiveness 32.11% 33.82%, WER, respectively. Code-mixed also accuracy up 37.74%