Intelligent System to Transformer Slang Words into Formal Words

Abstract

Understanding and utilizing informal words not recognized in standard dictionaries poses challenges for users. These words are specific to certain communities, hindering comprehension for those outside. Natural language processing tasks, like translation and summarization, struggle with informal vocabulary and local dialects. Although existing models can translate informal words, comprehensive solutions are elusive due to regional and contextual variations. Developing natural language processing models that consider informal words and local dialects is crucial for future research. This paper presents an updated dataset of informal English words tailored to current usage. Multiple models from the Transformer core library on the Hugging Face platform were trained and evaluated, with the facebook/bart-base model demonstrating high accuracy (training data loss: 0.05299). Continued research and innovation in this field are imperative for effective cross-cultural and intercommunity communication