From Long News to Accurate Forecast: Importance-Aware Fusion and PRM-Guided Reflection for Time Series Forecasting

In three linesNovel framework combining importance-aware news compression and process-level retrieval supervision for time series forecasting. A reward model estimates each article's forecasting utility for sequential fusion, while a PRM ranks supplementary-news candidates based on error profile. Experiments on finance, energy, traffic, and bitcoin benchmarks show improved accuracy and fewer refinement iterations.

Summary generated by Claude — human-verified