CogNarr

CogNarr is a soon-to-be-released, open-source, interactive tool that turns narrative into an explicit, formal semantic graph that you can inspect and correct until it faithfully reflects your beliefs.

People naturally express their beliefs about a topic, situation, or problem in the form of narrative. A CogNarr story graph captures what a narrative says, including what happens and when, to whom, how its statements relate as arguments, and whether each is asserted, predicted, or conditional. Where a word, a reference, or an identity is uncertain, the graph represents that uncertainty. In short, the story graph expresses the verifiable meaning of what a person wishes to say, how certain they are about it, and how certain the system is about its reading, all while reducing the ambiguity of natural language.

Applications

CogNarr has potential wherever consequential beliefs are expressed as narrative, from medicine, law, and governance to public polling and accident reporting, even customer returns. CogNarr story graphs also support accurate agent-agent communication. The collective-cognition aim behind the project is the alignment of shared belief models during decision-making in the large-group setting.

Invitation

If you can imagine using CogNarr in your work or research, we would like to hear from you. Collaboration, grants, and other forms of engagement or support are welcome. Relevant fields and topics include active inference, reinforcement learning, and artificial intelligence and computer science more broadly; cognitive science; computational semantics and linguistics; computational social science; applied category theory; knowledge representation; and medicine, governance, law, and collective decision-making.