Polish Press Agency

Knowledge‑graph‑based classification of market‑sensitive corporate disclosures

About the organisation

Polish Press Agency (PAP) is a national, public press agency. It provides other media, such as press, TV, radio and web portals with news (dispatches), photos and videos. Because of its public status it is also required to inform the public regarding national affairs. PAP’s mission is to deliver impartial, reliable and trustworthy data and information. Besides the general news platform, it also provides clients with content from specialised editorial offices. All of the products prepared by PAP have to be correctly described using metadata systems, so they can be correctly organised and distributed. 

What challenges can the GRAPHIA project help solve?

The Polish Press Agency fulfils its mandate of ensuring equal access to market information by publishing regulatory disclosures required from publicly listed companies. Publishing announcements from publicly listed companies must meet two conditions: the speed of delivery to the recipient (in financial markets, every piece of information is potentially price‑forming, and every second of delay creates information risk for investors) and the accuracy of message classification (assigning the report to the appropriate company and indicating the relevant corporate event facilitates the recipient’s investment decision‑making).

These processes must be fast and precise. A human can perform such classification accurately by understanding the context—but not quickly. Mechanisation based on searching for key phrases is unreliable in irregular situations.

What is the proposed use case?

The use of knowledge graphs can significantly support the realisation of this task by providing a structured, context‑aware representation of relationships between companies, corporate events, financial instruments, and regulatory concepts. By modelling entities and their connections, a knowledge graph enables automated systems to understand not only keywords, but also meaning and context, allowing reports to be accurately linked to the correct company and event type even in non‑standard or ambiguous cases.

In this use case, the knowledge graph serves as the foundation for an automatic classification mechanism, enabling fast and precise assignment of corporate communications to issuers and standardised event categories. It also supports an advanced search mechanism, allowing users to explore the content of corporate disclosures across long time horizons based on entities, event types, and semantic relationships rather than simple keyword matching. Furthermore, the knowledge graph can be shared with other institutions, including financial market participants, to harmonise information classification standards across the market. By opening access to this structured knowledge layer, the project creates opportunities for market participants to develop their own innovative data‑driven applications, strengthening transparency, interoperability, and analytical capabilities across the financial ecosystem.

What type of SSH data or content are involved?

As a national news agency, PAP operates both as a data provider for financial markets and as a publisher fulfilling a public mission of equal access to information. In both roles, they work primarily with structured and semi‑structured socio‑economic data related to corporate activity, market transparency, and investor behaviour.


As a data provider, PAP distributes regulated corporate communications originating from official and supervisory sources. This includes current and periodic reports of publicly listed companies, ad hoc disclosures, and announcements concerning financial performance, corporate actions, governance changes, and regulatory events. These materials are enriched with authoritative metadata, such as unique company identifiers and standardised taxonomies of corporate events, which are essential for professional investors and market infrastructure providers.

As a publisher, PAP focuses on making these communications widely available and discoverable, ensuring equal and timely access to price‑sensitive information. They seek to classify, contextualise, and semantically organise corporate announcements so they can be easily understood and compared by different audiences.

Additionally, PAP maintains knowledge organisation assets – taxonomies, identifiers, and concept hierarchies covering companies, events, and financial terminology – which constitute a key SSH resource supporting analysis of economic institutions, corporate behaviour, and information flows in financial markets.

Back to Industry Hub

Contact

Contact our GRAPHIA communication team via the email address:

contact@graphia-ssh.eu

Sign up for the GRAPHIA newsletter here.

Follow us on LinkedIn, Bluesky and YouTube.