dbbiyte/CISA-testset

CISA-testset是一个从İbrahim Temo的回忆录中提取的历史土耳其文本数据集，用于评估跨个体情感分析模型在真实历史文本上的性能。该数据集包含了200个句子，专注于历史人物和个性的情感分析，并具有连续性和真实性。数据集的创建旨在保持历史真实性，确保实体-情感关系的多样性，保留原始奥斯曼土耳其语的语言学特征，并提供全面的研究可再现性元数据。

CISA-testset is a historical Turkish text dataset extracted from İbrahim Temos memoirs, designed for evaluating the performance of Cross-Individual Sentiment Analysis models on real historical texts. The dataset includes 200 sentences, focusing on sentiment analysis of historical figures and personalities, with an emphasis on continuity and authenticity. The dataset was created to maintain historical authenticity, ensure diverse entity-sentiment relationships, preserve original Ottoman Turkish linguistic features, and provide comprehensive metadata for research reproducibility.