Search papers, labs, and topics across Lattice.
This paper introduces Information Terra, a novel semantic projection technique that visualizes document embeddings on a globe-like structure, with user-defined endpoints representing narrative anchors. By employing kernel density estimation to recover land features and constructing a narrative trail from a coherence graph, the method allows for an intuitive exploration of thematic and narrative progress across a document corpus. The approach is validated through a case study on the Cuban Protests, illustrating a coherent storyline from significant events over several years.
Visualizing document embeddings as a globe reveals narrative trajectories and thematic shifts in a way that traditional methods cannot match.
We introduce Information Terra, a narrative-anchored semantic-first projection that places a document corpus on an Earth-like globe whose poles are two user-chosen endpoint documents and whose prime meridian is the great-circle geodesic between them on the embedding hypersphere -- so latitude encodes narrative progress and longitude thematic deviation. Land features are recovered from document density via kernel density estimation and labeled by theme. A narrative trail built from the underlying narrative coherence graph, and constrained to be monotone in geodesic progress, provides a readable storyline. The projection's axes are semantically grounded in the user's chosen narrative endpoints, and the globe metaphor affords rotation and antipodal reading. We demonstrate the method on a 540-article Cuban Protests corpus, showing a storyline from Obama's 2016 visit to the 2021 International Aid during the protests.
