京东数码配件评价中英双语平行语料集

根据提供的数据说明书，该数据集名为“京东数码配件评价中英双语平行语料集”，是一个通过可控提示词工程、基于大语言模型批量虚拟生成的面向自然语言处理任务的中英双语平行语料集，共包含42242条有效数据。其内容模拟京东平台数码配件评价风格，覆盖78个品类及正面、负面、中性三种情感倾向，并具有用户画像、情感强度、传播力等30多个细粒度标注维度。According to the accompanying data specification, this dataset is named "JD Digital Accessories Review Bilingual Chinese-English Parallel Corpus". It is a Chinese-English bilingual parallel corpus for natural language processing tasks, which was virtually batch-generated by large language models (LLMs) through controllable prompt engineering, with a total of 42,242 valid entries. Its content simulates the review style of digital accessories on the JD.com platform, covering 78 product categories and three sentiment orientations: positive, negative and neutral. Additionally, it features more than 30 fine-grained annotation dimensions including user profile, sentiment intensity, dissemination capability and other relevant aspects.

搜集汇总

背景与挑战

背景概述

该数据集是一个通过大语言模型虚拟生成的中英双语平行语料集，包含42242条模拟京东数码配件评价的数据。它覆盖78个品类及正面、负面、中性三种情感倾向，并提供了30多个细粒度标注维度，适用于自然语言处理任务。