Overview#
An investigator monitoring a suspect company does not have time to scan forty news websites every morning. Nor can they rely on a keyword alert that fires on every mention of a common surname. The Argus News Intelligence Platform handles that monitoring automatically: ingesting articles from configured RSS feeds, analysing content for sentiment and entity relevance, correlating related stories across sources, and surfacing only what matters to active investigations. When a story breaks that references a person of interest, the platform has already linked it before the analyst opens their browser.
The platform includes bias detection, source credibility scoring, and AI-powered content expansion to help investigators understand the media landscape and identify coverage patterns across persons, organisations, and events under scrutiny.
Key Features#
- Multi-Source Aggregation: RSS feed integration with configurable polling intervals, manual article submission for paywalled sources, automatic deduplication, source domain tracking, and feed health monitoring.
- Content Analysis: AI-powered article expansion and summarisation, sentiment analysis, entity extraction for people, organisations, and locations, topic classification, and full-text search indexing.
- Source Credibility: Political bias tracking across the spectrum, credibility scoring, historical reputation tracking, and multi-source story verification for separating signal from noise.
- Story Correlation: Automatic grouping of related articles into stories with cross-source corroboration scoring, trending score calculation, timeline reconstruction, and source diversity analysis.
- Investigation Integration: Link articles to active investigations and profiles, generate investigation-specific trending scores, create keyword-based alerts, and correlate news events with investigation timelines.
Use Cases#
- Media Monitoring for Investigations: Configure keyword alerts to track news coverage related to active investigations, monitor emerging stories before they trend, and document media narrative timelines for prosecution or intelligence briefings.
- Source Verification: Assess credibility and bias of news sources by comparing coverage across politically diverse outlets, tracking credibility over time, and identifying potential disinformation campaigns.
- Entity Profile Enhancement: Enrich person and organisation profiles with news coverage through automatic entity linking, historical coverage timelines, and sentiment-based reputation tracking.
- Situational Awareness: Maintain awareness of breaking news and trending topics that may affect operations, with configurable source selection and relevance filtering tailored to operational areas of interest.
Integration#
The module connects with OSINT intelligence, investigation management, alert systems, and profile management modules within the platform. Administrators configure source management, bias filtering, credibility thresholds, and alert settings per tenant. All news data and entity links are stored in the PostgreSQL primary data store with organisation-level isolation.
Open Standards#
- RSS 2.0 / Atom (RFC 4287): Article ingestion is driven by polling news source feeds parsed with the
feedparserlibrary, which handles both RSS 2.0 and Atom 1.0 syndication formats; every monitored source carries anrss_feed_urland a configurable polling interval. - IPTC NewsCodes Media Topics: Topic classification uses the IPTC controlled vocabulary (
cv.iptc.org/newscodes/mediatopic/) with structured qcodes and scheme URIs, allowing articles and stories to be tagged with internationally recognised subject categories such as crime, terrorism, conflict, and disaster. - GraphQL (June 2018 specification): All client-facing queries, mutations, and subscriptions are exposed through a GraphQL API built with Strawberry, covering sources, articles, stories, entity links, alerts, and investigation-trending data.
- ISO 8601: Timestamps for article publication, feed fetch times, alert matches, and trending calculations are stored and exchanged in ISO 8601 format throughout the module.
- ISO 639-1 / BCP 47: A
language_codefield on both news sources and individual articles carries IETF language tags, enabling language-aware source filtering and content routing. - ISO 3166-1 alpha-2: A
country_codefield on news sources stores two-letter country codes, supporting geographic filtering and regional source discovery. - JSON Web Token (JWT, RFC 7519): The platform authenticates all GraphQL operations using JWT Bearer tokens issued by the central auth service, enforcing tenant-scoped access control across every news query and mutation.
Last Reviewed: 2026-02-05 Last Updated: 2026-04-14