🤖 AI Summary
This work proposes a semantics-first spherical projection method to construct interpretable, narrative-aware low-dimensional visualizations from high-dimensional document embeddings. By designating two user-specified documents as poles, the approach defines meridians via hyperspherical geodesics, where latitude encodes narrative progression and longitude captures thematic deviation. Kernel density estimation is integrated to generate terrain-like features and semantic labels on the sphere. The method introduces a novel narrative-anchor-driven spherical layout, ensuring monotonic story progression through a narrative coherence graph and supporting interactive rotation and antipodal reading. Evaluated on a corpus of 540 news reports covering Cuban protests, the framework successfully reconstructs a coherent narrative arc—from Obama’s 2016 visit to Cuba through to international aid efforts in 2021—demonstrating both its effectiveness and interpretability.
📝 Abstract
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.