Datasets

The four freely-available repositories from Webz.io each offer weekly snapshots of public-web news content, designed to support research, analytics and machine-learning applications. The Free News dataset provides roughly 1,000 articles per week drawn from a broad range of global news sites, with rich metadata including sentiment, categories and publication dates. The Political News dataset focuses on English-language political coverage (filtered via e.g., category:Politics language:english), allowing studies of political discourse and sentiment evolution. ([GitHub][2]) The Adverse Media News dataset concentrates on negative or high-risk media coverage within business/economy topics (e.g., site_type:news sentiment:negative category:"Economy, Business and Finance"), making it useful for compliance, risk or reputation monitoring use cases. Finally, the Financial News dataset isolates the finance & economic domain (via topic:"Financial and Economic News" language:english) to support models and analytics around markets, companies and financial events. Together they form a coherent suite of open resources for modelling and insight-generation across general, political, adverse-risk and financial news themes.