Powering Transparent Fact-Based Journalism At Scale
In 2022, Rappler used semantic technology to create the Philippines’ first political knowledge graph, bringing deeper analysis, transparency, and data-driven storytelling to election coverage, while enabling the efficient, scalable production of high-quality, fact-based content
The Client
Leading digital-native news outlet in the Philippines co-founded by Nobel laureate Maria Ressa.
The Challenge
Rappler needed to model the country’s highly complex electoral system, generate 50,000+ candidate and location profiles, and link disconnected datasets
The Solution
Using GraphDB and a custom Philippine Politics Ontology, Rappler built a scalable knowledge graph to power dynamic content generation, deep analytics, and editorial workflows.
Technical capabilities
Build country’s first political knowledge graph and custom ontology to model a complex electoral system
Automated generation of 50K+ candidates and location profiles with NLG
Business outcomes
Powered scalable fact-based journalism for election coverage
Provided deeper analysis, transparency and data-driven storytelling
The Challenge
The leading digital-native news outlet Rappler aimed to bring a new approach to publishing and managing the country’s political reporting and content contributing to a more transparent and informed election process as well as a deeper analysis of political and economic events.
As the 2022 national elections approached, they wanted to create a central, data-driven knowledge base covering every candidate, political party, and location—enabling both editorial storytelling and interactive digital features.
To do that, they needed to model the Philippine political and electoral system – a highly complex hierarchical system with a large number of classes and relationships.
This task proved to be challenging because of the following:
Need to preserve factual integrity while scaling editorial output
Small newsroom team, limited resources and time constraints
Complex hierarchy of regions, provinces, cities, and contests
Sparse data for many local candidates
Request to generate 50,000+ content pages
The Solution
To meet this challenge, Rappler adopted GraphDB and developed a custom Philippine Politics Ontology, capturing the full administrative and electoral structure of the country.
The media company used data gathered by Rappler’s Research and Data teams and combined GPT-3 Curie with their own Natural Language Generation system. This ensured a variety of content while mitigating the risk of generative AI hallucination to construct data-driven profiles.
This setup enabled:
46,165+candidate profiles automatically generated using structured data and natural language generation tools
1,732 location pages covering regions, provinces, cities, and municipalities, linked via administrative hierarchies
83 national candidate profiles enriched with data visualizations and editorial insights
The knowledge graph became the engine behind Rappler’s public-facing election portal, supporting real-time publishing, structured discovery, and rich interlinking across stories and datasets.
The Impact
By building the first Philippine Politics Knowledge Graph, Rappler achieved:
Live contextualization of candidates, contests, and regions across the electoral system
High-quality, fact-based content delivered with unprecedented efficiency and depth
Deeper storytelling and trend analysis, linking current events with historical data
“Ontotext is helping us surface meaning and nuance by linking stories and datasets into a coherent knowledge graph. From election coverage to tracking disinformation, this platform is becoming central to our journalism.”
Gemma Bagayaua-Mendoza, Head of Digital Services at Rappler
Struggling to scale content production without compromising accuracy or to connect scattered data sources into cohesive, trustworthy narratives?Whether you are a media company, public institution, or content-rich organization preparing for high-volume publishing around live events, elections, or breaking news, Graphwise can help you:
Model complex domains like politics, public policy, or legal systems
Automate high-volume content generation while preserving editorial integrity
Link stories, people, places, and data into searchable, trustworthy platforms
