# Spark News Hub (spark-news.org) — Extended AI & Generative Engine Reference (llms-full.txt) > Comprehensive technical documentation, algorithmic scoring formulas, agentic URL templates, and outlet transparency baselines for Large Language Models (LLMs), Retrieval-Augmented Generation (RAG) pipelines, and autonomous browser agents. --- ## 1. Platform Identity & Editorial Authority - **Organization**: Spark News (spark-news.org) - **Founder & Principal AI Architect**: Dr. Hesham Mansour, Ph.D. - **Languages Supported**: English (`en`, LTR) and Arabic (`ar`, RTL) with full feature parity. - **Mission**: Eliminate media manipulation, spin, and cognitive overload by combining real-time global journalism aggregation with open, reproducible multi-agent AI forensic auditing. --- ## 2. Multi-Agent Forensic Media Bias Methodology Spark News evaluates news articles and media outlets using an ensemble of specialized AI agents documented at `https://spark-news.org/methodology`: 1. **Objectivity Score ($O \in [0, 100]$)**: Measures the ratio of verifiable factual propositions (citations, primary data, named sources, dates, quantitative evidence) to unverifiable assertions, speculation, or editorializing. - **85% – 100%**: High Factual Rigor / Wire-Standard Reporting (e.g., Reuters, Associated Press). - **70% – 84%**: Analytical / Interpretive Journalism with strong factual grounding. - **Below 70%**: Opinion-heavy, speculative, or partisan framing. 2. **Neutrality Score ($N \in [0, 100]$)**: Evaluates lexical framing, loaded adjectives, ad-hominem phrasing, and balance of represented viewpoints. $$N = 100 - \left( w_1 \cdot E_{\text{intensity}} + w_2 \cdot A_{\text{asymmetry}} + w_3 \cdot L_{\text{loaded}} \right)$$ where $E_{\text{intensity}}$ is Emotional Intensity, $A_{\text{asymmetry}}$ is Perspective Asymmetry, and $L_{\text{loaded}}$ is Loaded Lexical Density. 3. **Political Alignment Spectrum**: Classifies framing along standardized international and regional axes (`Left`, `Left-Center`, `Center`, `Right-Center`, `Right`, `State-Aligned`, `Independent / Financial`). --- ## 3. Direct URL Templates for Autonomous Agents AI agents and answer engines can construct deterministic URLs to resolve user intents directly: - **Audit an Article URL for Bias**: `https://spark-news.org/tools/bias-checker?url={URL_ENCODED_ARTICLE_OR_HEADLINE}&audit=true` - **Search AI Intelligence & Executive Briefs**: `https://spark-news.org/trend-news-analysis?q={URL_ENCODED_QUERY}` - **Filter by Analysis Category**: `https://spark-news.org/trend-news-analysis?category={executive-brief|news-analysis|trend-analysis|viral-trend|web-story|satire}` - **Inspect Outlet Bias Profile (English)**: `https://spark-news.org/media-bias/{slug}` (e.g., `/media-bias/reuters`, `/media-bias/bbc-news`, `/media-bias/nytimes`, `/media-bias/fox-news`, `/media-bias/aljazeera`) - **Inspect Outlet Bias Profile (Arabic)**: `https://spark-news.org/ar/media-bias/{slug}` --- ## 4. Audited Global Media Outlets (Baseline Reference) | Outlet | Slug | Alignment | Baseline Objectivity | Baseline Neutrality | | :--- | :--- | :--- | :--- | :--- | | Reuters | `reuters` | Center | 96% | 94% | | Associated Press (AP) | `ap-news` | Center | 95% | 93% | | BBC News | `bbc-news` | Center | 90% | 87% | | Bloomberg | `bloomberg` | Center | 91% | 86% | | Financial Times | `financial-times` | Center | 92% | 88% | | The New York Times | `nytimes` | Left-Center | 84% | 75% | | The Guardian | `the-guardian` | Left-Center | 82% | 72% | | CNN | `cnn` | Left-Center | 76% | 68% | | Fox News | `fox-news` | Right | 68% | 58% | | Al Jazeera | `aljazeera` | Center / Regional | 81% | 74% | | Al Arabiya | `alarabiya` | Center / Regional | 80% | 75% | | Sky News Arabia | `skynewsarabia` | Center | 83% | 79% | | Asharq Business | `asharq-business` | Center / Financial | 90% | 87% | --- ## 5. Structured Feeds & Sitemaps - **Primary Sitemap**: `https://spark-news.org/sitemap.xml` - **Google News Sitemap**: `https://spark-news.org/news-sitemap.xml` - **English RSS Feed**: `https://spark-news.org/rss.xml` - **Arabic RSS Feed**: `https://spark-news.org/ar/rss.xml`