淘宝网笔记本电脑评价中英双语平行语料集

本数据集为"淘宝网笔记本电脑商品评价双语语录集"（v1.0），由百图科技工作室基于大语言模型通过可控提示词工程批量虚拟生成。数据集模拟了淘宝网笔记本电脑商品评价的真实风格，共包含100,000条有效双语语录，每条语录均提供简体中文与地道英文对照版本，中文长度控制在2-120字，英文控制在1-80词。数据集配备语录属性标注、用户画像模拟、适用场景标注、传播力模拟、情感强度分级、翻译质量评估等六大高价值细粒度标注维度，所有标注置信度均为1.0。全部数据均为虚拟生成，不涉及任何真实用户隐私，无需脱敏处理，可安全用于各类自然语言处理研究与模型训练任务。This dataset is the *Bilingual Quotes Collection of Taobao Laptop Product Reviews* (v1.0), which was batch-generated virtually by Baitu Technology Studio using Large Language Models (LLMs) via controllable prompt engineering. It simulates the authentic style of real Taobao laptop product reviews, containing a total of 100,000 valid bilingual quotes. Each quote provides a parallel pair of Simplified Chinese and idiomatic English, with the length of Chinese content restricted to 2–120 characters and English content restricted to 1–80 words. The dataset includes six high-value fine-grained annotation dimensions: quote attribute annotation, simulated user profiles, applicable scenario annotation, spreadability simulation, emotional intensity grading, and translation quality evaluation, with all annotation confidence scores set to 1.0. All data in this dataset are virtually generated, without involving any real user privacy and requiring no de-identification processing, making it safe for various natural language processing (NLP) research and model training tasks.

搜集汇总

背景与挑战

背景概述

本数据集是一个由百图科技工作室通过大语言模型虚拟生成的淘宝网笔记本电脑评价中英双语平行语料集，包含10万条模拟真实评价风格的中文与英文对照语录。数据集提供了语录属性、用户画像、适用场景、传播力、情感强度和翻译质量等六大细粒度标注，专为自然语言处理研究任务设计，且不涉及真实用户隐私。