京东家电评价中英双语平行语料集

本数据集模拟京东平台家电评价内容风格，涵盖100种家电品类、1000个商品关键词，由大语言模型虚拟生成100,000条中英双语平行语料。每条语料包含中文评价及对应地道英文翻译，附有语料属性、用户画像、适用场景、传播力、情感强度、翻译质量等六大高价值标注维度。全部虚拟生成，无隐私风险。This dataset simulates the review style of home appliance products on the JD.com platform, covering 100 home appliance categories and 1,000 product keywords. It contains 100,000 sets of Chinese-English parallel corpora virtually generated by Large Language Models (LLMs). Each corpus includes a Chinese review and its corresponding authentic English translation, with six high-value annotation dimensions: corpus attributes, user profiles, applicable scenarios, communication potential, emotional intensity, and translation quality. All content is virtually generated and poses no privacy risks.

搜集汇总

背景与挑战

背景概述

该数据集是一个由大语言模型虚拟生成的中英双语平行语料集，模拟了京东平台的家电评价风格，包含10万条语料，覆盖100种家电品类和1000个商品关键词。每条语料均提供中文评价及地道英文翻译，并附有语料属性、用户画像、适用场景、传播力、情感强度和翻译质量六大高价值标注维度，全部为虚拟生成内容，无隐私风险。