Scholarlink

This article describes a comprehensive approach to text analysis that combines several modern natural language processing models, including spaCy, Word2Vec, BERT, and ruGPT-3. The primary goal of the development was to create a method capable of considering the contextual aspects of text, which is critical for accurate analysis of short messages on social media. The method uses multi-level analysis to achieve high accuracy in classifying texts by emotional content, taking into account the contextual significance of words and the overall emotional direction of phrases. This approach not only enhances accuracy in detecting subtle emotional cues but also adapts to the brief and often ambiguous nature of social media communication.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Unveiling Urban Emotions: Contextual Alignment in Sentiment Analysis of Social Media Microtexts

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Anna Chizhik,
  • Katarina Cokrlic

摘要

This article describes a comprehensive approach to text analysis that combines several modern natural language processing models, including spaCy, Word2Vec, BERT, and ruGPT-3. The primary goal of the development was to create a method capable of considering the contextual aspects of text, which is critical for accurate analysis of short messages on social media. The method uses multi-level analysis to achieve high accuracy in classifying texts by emotional content, taking into account the contextual significance of words and the overall emotional direction of phrases. This approach not only enhances accuracy in detecting subtle emotional cues but also adapts to the brief and often ambiguous nature of social media communication.