Gaining an overview and being able to respond to current political issues as they unfold requires digital assistance. Whether it’s about scanning documents for mentions of relevant keywords or about coordinating activities across teams, technology is your friend.

Applied graph theory is our key to unleashing network effects

The core of Ulobby's data model is a knowledge graph that encompasses a representation of categories, properties and relations between the concepts, data and entities within political domains of discourse.

As today's digital processes create more and more data, data silos often inhibit efficient information usage and seamless digital processes. The knowledge graph enables Ulobby to connect existing data sources and systems, creating new insights and knowledge by semantically enriching information and data flows. Having a connected body of information allows us to ask new questions and discover new answers within the existing data. In policy processes the relationships between data points often matter more than the individual points themselves, and by mapping all data including stakeholders and stakeholder relations in a directed mathematical structure Ulobby allows for complex inferences of relevant stakeholders and their positions in relation to your policy issues. For an effective public affairs organization, the political network is a living asset that is cultivated and developed.

Knowledge extraction from unstructured texts

Modern democracies produce an incomprehensible number of texts. The news cycle that was once dictated by the limitations of print media distribution is now a constant stream across a plethora of channels.

Translating texts from prose form to data that machines can process is a science that we cultivate religiously at Ulobby through ongoing research collaborations with universities and GTS institutes.Reading and decoding political positions and policy documents is difficult even for humans, but recent technological developments within the field of Natural Language Processing have allowed our team to develop and apply a selection of algorithms that make the work much easier. For example, we use topic recognition to discover and follow the agendas relevant to our users, named-entity recognition to determine the stakeholders in a given debate, and opinion mining to explore arguments and sentiments within a given political issue.