地球资源数据云——数据资源详情
多标签分类数据集
该数据集《Multi - Label Classification Dataset》主要用于多分类任务，数据形态以文本为主，应用场景偏向金融风控。 题目说明：Topic Modeling for Research Articles 任务类型：文本多分类。 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：sample_submission.csv, test.csv, train.csv。 Context NLP: Multi - label Classification Dataset. Content The dataset contains 6 different labels(Computer Science, Physics, Mathematics, Statistics, Quantitative Biology, Quantitative Finance) to classify the research papers based on Abstract and Title. The value 1 in label columns represents that label belongs to that paper. Each paper has multiple labels as 1. Acknowledgements
摘要概览
该数据集《Multi - Label Classification Dataset》主要用于多分类任务，数据形态以文本为主，应用场景偏向金融风控。 题目说明：Topic Modeling for Research Articles
任务类型：文本多分类。
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：sample_submission.csv, test.csv, train.csv。
Context
NLP: Multi - label Classification Dataset.
Content
The dataset contains 6 different labels(Computer Science, Physics, Mathematics, Statistics, Quantitative Biology, Quantitative Finance) to classify the research papers based on Abstract and Title. The value 1 in label columns represents that label belongs to that paper. Each paper has multiple labels as 1.
Acknowledgements