NUBES

NUBES是由维科科技基金会创建的西班牙语生物医学文本数据集，专注于否定和不确定性标注。该数据集包含29,682个从匿名健康记录中提取的句子，这些句子均被标注了否定和不确定性信息。NUBES是目前公开的最大西班牙语否定标注数据集，也是首个包含推测线索、范围和事件标注的数据集。数据集的创建过程涉及从医院获取匿名健康记录，然后通过人工和半自动方法进行个人信息去标识化处理。NUBES的应用领域主要集中在生物医学领域，旨在通过捕捉文本中的否定和不确定性信息，帮助区分疾病状态或药物使用情况，从而提高自然语言理解的准确性。

NUBES is a Spanish biomedical text dataset developed by the Vico Technology Foundation, dedicated to negation and uncertainty annotation. It contains 29,682 sentences extracted from anonymized health records, all of which have been annotated with negation and uncertainty information. To date, NUBES is the largest publicly available Spanish dataset for negation annotation, and it is also the first dataset to include annotations for speculative cues, scopes and events. The dataset's development process entails obtaining anonymized health records from hospitals, followed by manual and semi-automated de-identification of personal information. Its primary applications are in the biomedical domain, where it aims to help distinguish disease states or medication usage by capturing negation and uncertainty information in texts, thereby enhancing the accuracy of natural language understanding.