ORACLE: Time-Dependent Recursive Summary Graphs for Foresight on News Data Using LLMs
This work addresses the need for automated, domain-specific news analysis to support curriculum intelligence, but it is incremental as it applies existing methods like clustering and LLMs to a new dataset.
The authors tackled the problem of generating timely, decision-ready insights from daily news for a Finnish University of Applied Sciences by developing ORACLE, a platform that processes news into weekly summaries using clustering and LLMs, resulting in a system that highlights changes and groups them into themes for analysis.
ORACLE turns daily news into week-over-week, decision-ready insights for one of the Finnish University of Applied Sciences. The platform crawls and versions news, applies University-specific relevance filtering, embeds content, classifies items into PESTEL dimensions and builds a concise Time-Dependent Recursive Summary Graph (TRSG): two clustering layers summarized by an LLM and recomputed weekly. A lightweight change detector highlights what is new, removed or changed, then groups differences into themes for PESTEL-aware analysis. We detail the pipeline, discuss concrete design choices that make the system stable in production and present a curriculum-intelligence use case with an evaluation plan.