地球资源数据云——数据资源详情
灾难推文
该数据集《Disaster Tweets》主要用于监督学习任务，数据形态以文本为主，应用场景偏向文本内容分析。 题目说明：Real or Not? NLP with Disaster Tweets challenge add - on 任务类型：文本监督学习。 建议流程：先做文本清洗与分词，再比较 TF - IDF+线性模型 与 预训练语言模型。 评估建议：使用分层切分或交叉验证，优先关注 F1、Recall、AUC 等分类指标。 可用文件：tweets.csv。 Context The file contains over 11,000 tweets associated with disaster keywords like “crash”, “quarantine”, and “bush fires” as well as the location and keyword itself. The data structure was inherited from Disasters on social media The tweets were collected on Jan 14th, 2020. Some of the topics people were tweeting: The eruption of Taal Volcano in Batangas, Philippines
摘要概览
该数据集《Disaster Tweets》主要用于监督学习任务，数据形态以文本为主，应用场景偏向文本内容分析。 题目说明：Real or Not? NLP with Disaster Tweets challenge add - on
任务类型：文本监督学习。
建议流程：先做文本清洗与分词，再比较 TF - IDF+线性模型 与 预训练语言模型。
评估建议：使用分层切分或交叉验证，优先关注 F1、Recall、AUC 等分类指标。
可用文件：tweets.csv。
Context
The file contains over 11,000 tweets associated with disaster keywords like “crash”, “quarantine”, and “bush fires” as well as the location and keyword itself. The data structure was inherited from Disasters on social media
The tweets were collected on Jan 14th, 2020.
Some of the topics people were tweeting:
The eruption of Taal Volcano in Batangas, Philippines