地球资源数据云——数据资源详情
亚马逊 Kosmos 数据集
该数据集《Amazon Kosmos Dataset》主要用于多分类任务，数据形态以文本为主，应用场景偏向文本内容分析。 题目说明：NLP - Sentiment Analysis, Customer Comments 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：kozmos.csv。 Cosmos, which produces home textiles and daily wear focused products selling on Amazon, increases its sales by analyzing the comments made on its products and improving its features according to the complaints it receives. aims to increase. For this purpose, comments will be labeled by performing sentiment analysis and a classification model can be created with the labeled data. Star: Number of stars given to the product HelpFul: Number of people who found the comment helpful Title: Title given to comment content, short comment
摘要概览
该数据集《Amazon Kosmos Dataset》主要用于多分类任务，数据形态以文本为主，应用场景偏向文本内容分析。 题目说明：NLP - Sentiment Analysis, Customer Comments
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：kozmos.csv。
Cosmos, which produces home textiles and daily wear focused products selling on Amazon, increases its sales by analyzing the comments made on its products and improving its features according to the complaints it receives. aims to increase.
For this purpose, comments will be labeled by performing sentiment analysis and a classification model can be created with the labeled data.
Star: Number of stars given to the product
HelpFul: Number of people who found the comment helpful
Title: Title given to comment content, short comment