Browsing by Author "Namazbayev A."

Now showing 1 - 1 of 1
  • Thumbnail Image
    ItemOpen Access
    NLP-Based Offensive Language Detection in Text Messages
    (2024) Namazbayev A.
    This dissertation explores the enhancement of three Natural Language Processing (NLP) models—BERT, FastText, and LSTM—through integration with MetaLlama-3-8B-Finetuned for detecting offensive language within text messages. This integration aims to improve the precision, recall, and overall effectiveness of these models in moderating digital communications. By assessing each model’s performance before and after integration, this study highlights the significant gains in their ability to handle complex linguistic patterns and provides a comparative analysis of their enhanced capabilities.