ICIS Chemical Business Archive (1876–1930)
收藏资源简介：
Complete pre-1930 archive of ICIS Chemical Business, the leading trade journal for the petroleum, paint, and chemical industries. **763,738 rows** of clean, structured text documenting industrial markets from 1876 to 1930. **What this data does for your model:** - Your model learns authentic 19th‑ and early 20th‑century chemical industry reporting from one of the oldest continuously published trade journals. - Your model retrieves original market intelligence on petroleum pricing, oil refining, paint and pigment trades, and naval stores (turpentine, resin). - Your model trains on the language of industrial commodities, supply chain logistics, and chemical manufacturing during the rise of the modern oil industry. - Your model understands the evolution of chemical market reporting, from wild rubber auctions to plantation grades and the Stevenson Restriction Plan. **What's inside:** - Petroleum pricing and oil market reports - Chemical trade news and commodity prices - Paint, varnish, and pigment industry intelligence - Naval stores, turpentine, and resin markets - Industrial development in the age of oil **Perfect for:** - LLM fine-tuning on chemical industry text - Commodity trading and market analysis - Economic history and industrial development research - Energy and petroleum industry intelligence **Format:** Snowflake-native JSONL with columns: ISSUE, TITLE, AUTHOR, TYPE, TEXT. Fully cleaned, bias-audited, and ready for AI training. *From the birth of the petroleum age, chemical market intelligence, now ready for AI.* ***Cortex Agent Prompts:*** 1. Synthesize the industrial reporting and market analysis from this archive to track the evolution of global chemical supply chains and commercial production standards throughout the 20th century. 2. Evaluate historical patterns in chemical manufacturing innovation, regulatory milestones, and resource management strategies documented within this corpus to support predictive modeling and trend analysis in modern industrial chemistry. 3. Extract and analyze key developments in chemical industry trade policies and corporate reporting practices from this dataset to create foundational RAG models for historical and economic research in the chemical sector. <p><br/></p>
ICIS Chemical Business Archive (1876–1930) 数据集概述
数据集基本信息
- 数据集名称：ICIS Chemical Business Archive (1876–1930)
- 提供商：Devin Media Corp.
- 数据量：763,738 行
- 数据格式：Snowflake-native JSONL
- 数据覆盖时间范围：1876 年至 1930 年
- 数据内容：清洁、结构化的文本，记录了 1876 年至 1930 年的工业市场
- 数据主题：石油、涂料和化学工业的领先贸易期刊完整档案
数据内容详情
主要内容涵盖
- 石油定价和石油市场报告
- 化学品贸易新闻和商品价格
- 涂料、清漆和颜料行业情报
- 松脂制品、松节油和树脂市场
- 石油时代的工业发展
数据表结构
- 表名：ICIS_CORPUS
- 列信息：
- ISSUE (Varchar)
- TITLE (Varchar)
- AUTHOR (Varchar)
- TYPE (Varchar)
- TEXT (Varchar)
- INGESTION_DATE (Timestamp_NTZ)
数据预处理
- 完全清洁
- 偏见审核
- 适用于 AI 训练
适用场景
- 针对化学行业文本的 LLM 微调
- 商品交易和市场分析
- 经济史和工业发展研究
- 能源和石油行业情报
技术应用
机器学习
- 训练、微调和部署机器学习模型
- 适用于领域特定的 LLM 微调、商品市场术语提取和工业 NLP
真实世界数据 (RWD)
- 利用历史记录的石油定价、化学品市场和工业商品作为研究和分析的真实世界数据
- 捕捉现代石油工业的诞生
生命科学商业化
- 支持化学和能源研究，提供记录石油精炼、化学品制造和工业市场演变的精选历史行业文献
数据偏见审核说明
- 数据经过 Foundation for Ethical AI 的偏见审核
- 文本可能包含反映其时代种族、性别、文化和社会偏见的语言、术语和观点
- 此类语言不代表当代标准、价值观或医学理解
- 旨在实现历史知情的 AI 开发，而非延续有害语言、意识形态或任何形式的歧视
数据更新与可用性
- 更新频率：每年
- 地理覆盖范围：全球
- 云区域可用性：
- AWS：加拿大（中部）、欧盟（法兰克福）、欧盟（爱尔兰）、欧盟（伦敦）等 31 个区域
定价与获取
- 定价模式：按需请求
- 数据产品：化学工业遗产收藏
提供商其他数据集
- The Canadian Nurse Archive (Pre-1930)
- JAMA Surgery Dataset 1911 to 1930 Cleaned Structured AI Ready
- American Journal of Orthodontics (1915-1930)
- Medical Journal of Australia Archive (1914-1930)
- JAMA Pediatrics Clinical Dataset (1911–1930)
- Canadian Journal of Public Health Archive (1913-1930)
