ORACLE: Time-Dependent Recursive Summary Graphs for Foresight on News Data Using LLMs

TL;DR

ORACLE is a system that processes daily news to generate weekly, decision-ready insights using LLMs, by embedding, classifying, and summarizing news content into a dynamic, PESTEL-aware graph structure for analysis.

Contribution

The paper introduces a novel pipeline that creates a time-dependent recursive summary graph from news data, integrating change detection and PESTEL classification for decision support.

Findings

Effective weekly news summarization and classification.

Stable production pipeline with change detection.

Use case demonstrating curriculum intelligence insights.

Abstract

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.