Fortress: A Case Study in Stabilizing Search Recommendations via Temporal Data Augmentation and Feature Pruning
The paper introduces Fortress, a framework designed to stabilize search recommendations by addressing temporal instability in predictive models. It focuses on identifying and pruning features that cause inconsistent predictions, thereby enhancing model reliability. Fortress has been validated through experiments, showing significant improvements in prediction stability and classification performance.
- ▪Fortress aims to enhance model stability and accuracy in search and recommendation systems.
- ▪The framework follows a four-step process to prune instability-inducing features and retrain models.
- ▪Experiments demonstrate notable improvements in prediction stability and classification performance.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Information Retrieval arXiv:2605.15299 (cs) [Submitted on 14 May 2026] Title:Fortress: A Case Study in Stabilizing Search Recommendations via Temporal Data Augmentation and Feature Pruning Authors:Milind Pandurang Jagre, Jia Huang, Dayvid V. R. Oliveira, Zhinan Cheng, Babak Seyed Aghazadeh, Puja Das, Chris Alvino, Jinda Han, Kailash Thiyagarajan View a PDF of the paper titled Fortress: A Case Study in Stabilizing Search Recommendations via Temporal Data Augmentation and Feature Pruning, by Milind Pandurang Jagre and 8 other authors View PDF HTML (experimental) Abstract:In search and recommendation systems, predictive models often suffer from temporal instability when certain input features introduce volatility in output scores.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
