مشروع البحث:
A RULE-BASED NORMALIZATION MODEL WITH EMBEDDING TECHNIQUES FOR BI-LSTM HATE SPEECH DETECTION MODEL

thumbnail.project.alt
المساهمين
الممولين
رقم التعريف
404
الباحث
ZAINAB A ALSAIED MANSUR
المشرفين
منشورات
وحدات تنظيمية
thumbnail.orgunit.alt
وحدة تنظيمية
الوصف
ABSTRACT The topic of identifying hate speech at scale remains unresolved despite the several strategies lately proposed in natural language processing research for detecting various types of abusive language. The challenge facing the online hate speech detection problem is that the content generated by users resorts to being noisy that may contain shorter slang words or different creative spelling. However, normalization methods have not been extensively examined in the hateful context to reduce the out-of￾vocabulary (OOV) issue. Among the words used to compose social media messages are words that have repeated letters. Several techniques have been developed to normalize words with repeated letters. However
الكلمات الدالة
A RULE-BASED NORMALIZATION MODEL WITH EMBEDDING TECHNIQUES FOR BI-LSTM HATE SPEECH DETECTION MODEL