土耳其语目标情感分析数据集

本数据集由博阿齐奇大学创建，专门用于土耳其语的目标情感分析，包含3952条推文，覆盖2020年1月至6月的时间段。数据集通过Twitter API收集，针对6个知名公司和品牌进行筛选。每条推文均由两位土耳其语母语者手动标注，分别标注整体句子和针对特定目标的情感倾向。数据集主要用于解决传统情感分析无法区分文本中不同目标情感的问题，适用于开发和评估针对特定目标的情感分析模型。

This dataset was created by Boğaziçi University specifically for Turkish language target-based sentiment analysis. It contains 3,952 tweets spanning the period from January to June 2020. The dataset was collected via the Twitter API and filtered for 6 well-known companies and brands. Each tweet was manually annotated by two native Turkish speakers, with sentiment polarity labeled for both the entire sentence and specific target entities. This dataset is primarily designed to address the limitation of traditional sentiment analysis that fails to distinguish sentiment towards different targets in text, and is suitable for developing and evaluating target-based sentiment analysis models.