Hey r\/datasets<\/a> community!<\/p>\n I\u2019m thrilled to share an exciting new resource for all you data enthusiasts, researchers, and finance aficionados out there. https:\/\/github.com\/sovai-research\/open-investment-datasets<\/a><\/p>\n \ud83d\udd0d What\u2019s New?<\/strong><\/p>\n Sov.ai<\/a> has just launched the Open Investment Data Initiative<\/strong>! We\u2019re building the industry\u2019s first open-source investment datasets tailored for rigorous research and innovative projects. Whether you’re into AI, ML, quantitative finance, or just love diving deep into financial data, this is for you.<\/p>\n \ud83d\udcc5 Free Access with a 6-Month Lag<\/strong><\/p>\n All our 20 datasets<\/strong> will be available for free<\/strong> with a 6-month lag<\/strong> for non-commercial research purposes. This means you can access high-quality, ticker-linked data without breaking the bank. For commercial use, we offer a subscription plan that makes premium data affordable (more on that below).<\/p>\n \ud83d\udcc8 What We Offer<\/strong><\/p>\n By the end of 2026, Sov.ai<\/a> aims to provide 100+ investment datasets<\/strong>, including but not limited to:<\/p>\n \ud83d\udcf0 News Sentiment:<\/strong> Ticker-matched and theme-matched sentiment analysis from various news sources. \ud83d\udcc8 Price Breakout Predictions:<\/strong> Daily updates predicting upward price movements for US equities. \ud83d\udd0d Insider Flow Prediction:<\/strong> Over 60 insider trading features ideal for machine learning models. \ud83d\udcbc Institutional Trading:<\/strong> In-depth analysis of institutional investment behaviors and strategies. \ud83d\udce2 Lobbying Data:<\/strong> Detailed data on corporate lobbying activities, linked to specific tickers. \ud83d\udc8a Pharma Clinical Trials:<\/strong> Unique dataset tagging clinical trials with predicted success outcomes. \u26a0\ufe0f Corporate Risks:<\/strong> Bankruptcy predictions (Chapter 7 & 11) for over 13,000 US publicly traded stocks. …and many more! <\/p>\n \ud83e\udd1d Get Involved!<\/strong><\/p>\n We\u2019re looking for firms and individuals to join us as co-architects or sponsors on this journey. Your support can help us expand our offerings and maintain the quality of our data. Interested? Reach out to us here<\/a> or connect via our LinkedIn<\/a>, GitHub<\/a>, and Hugging Face<\/a> profiles.<\/p>\n \ud83e\uddea Example Use Cases<\/strong><\/p>\n Here\u2019s how easy it is to get started with our datasets using the Hugging Face datasets library:<\/p>\n from datasets import load_dataset<\/p>\n # Example: Load News Sentiment Dataset<\/p>\n df_news_sentiment = load_dataset(“sovai\/news_sentiment”, split=”train”).to_pandas()<\/p>\n # Example: Load Price Breakout Dataset<\/p>\n df_price_breakout = load_dataset(“sovai\/price_breakout”, split=”train”).to_pandas()<\/p>\n # Add more datasets as needed…<\/p>\n<\/div>\n submitted by \/u\/OppositeMidnight <\/a>
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
