/memetics-research-summaries

Memetics Reserach Summaries

Contents

Papers

Generation and Creation

Applications and Special Uses

Safety and Robustness

Detection and Classification

Evolution of meme classification approaches:
From building custom classifiers to leveraging pre-trained models.

  1. Pre-Embedding Era:
    • Custom architectures (MemeFier, MemeTector)
    • Task-specific models
    • Heavy feature engineering
  2. Modern CLIP/Embedding Era:
    • Leveraging pre-trained models
    • Focus on adapting large-scale embeddings
    • Multiple CLIP-based approaches
  3. Specialized approaches remain relevant for:
    • Limited resources
    • Domain-specific needs
    • Interpretability
    • Specific performance requirements

Emotion and Sentiment Analysis

Hate and Toxicity Detection

Misogyny Identification

Analysis and Understanding

Virality and Evolution Studies

Linguistic and Content Analysis

Financial Impact

Datasets and Tools

Books

Defender's Corner: Culture Science