Oracle : Time-Dependent Recursive Summary Graphs for Foresight on News Data Using LLMs
Lev Kharlashkin, Eiaki Morooka, Yehor Tereshchenko, Mika Hämäläinen · 2025
Read Oracle : Time-Dependent Recursive Summary Graphs for Foresight on News Data Using LLMs by Lev Kharlashkin, Eiaki Morooka, Yehor Tereshchenko, Mika Hämäläinen (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-15397.txt.
Description
Lev Kharlashkin and three coauthors build ORACLE, a platform that turns daily news into foresight for a Finnish university of applied sciences. Leaders need week over week, decision ready views of the world, sorted by the PESTEL lenses of politics, economy, society, technology, environment, and law. Raw news volume makes that impossible by hand.
ORACLE crawls and versions news, filters it for the university's relevance, embeds and classifies each item into PESTEL dimensions. It then builds a time dependent recursive summary graph: two layers of clusters, each summarized by a language model, rebuilt weekly. A change detector flags what is new, removed, or altered, and groups the differences into themes. The paper details the pipeline and the design choices that keep it stable in production, and it presents a curriculum intelligence use case with an evaluation plan.
Catalog record
| CREATOR | Lev Kharlashkin, Eiaki Morooka, Yehor Tereshchenko, Mika Hämäläinen |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 14 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-15397