# Optimizing Feature Extraction for On-device Model Inference with User Behavior Sequences

arXiv: https://arxiv.org/abs/2603.21508 · HTML: https://mlsanity.com/paper/2603.21508/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 79 references that are in the map.
- First posted on arXiv 23 Mar 2026.

## Builds on (most cited first, 21 of 21)

- [AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration](https://mlsanity.com/paper/2306.00978/) (2023)
- [Wide & Deep Learning for Recommender Systems](https://mlsanity.com/paper/1606.07792/) (2016)
- [Data-centric Artificial Intelligence: A Survey](https://mlsanity.com/paper/2303.10158/) (2023)
- [PatDNN: Achieving Real-Time DNN Execution on Mobile Devices with Pattern-based Weight Pruning](https://mlsanity.com/paper/2001.00138/) (2020)
- [DNNFusion: Accelerating Deep Neural Networks Execution with Advanced Operator Fusion](https://mlsanity.com/paper/2108.13342/) (2021)
- [Feature Selection: A Data Perspective](https://mlsanity.com/paper/1601.07996/) (2016)
- [PCONV: The Missing but Desirable Sparsity in DNN Weight Pruning for Real-time Execution on Mobile Devices](https://mlsanity.com/paper/1909.05073/) (2019)
- [A First Look at Deep Learning Apps on Smartphones](https://mlsanity.com/paper/1812.05448/) (2018)
- [PruneTrain: Fast Neural Network Training by Dynamic Sparse Model Reconfiguration](https://mlsanity.com/paper/1901.09290/) (2019)
- [Mandheling: Mixed-Precision On-Device DNN Training with DSP Offloading](https://mlsanity.com/paper/2206.07509/) (2022)
- [Smart at what cost? Characterising Mobile Deep Neural Networks in the wild](https://mlsanity.com/paper/2109.13963/) (2021)
- [On Application of Learning to Rank for E-Commerce Search](https://mlsanity.com/paper/1903.04263/) (2019)
- [LUT-NN: Empower Efficient Neural Network Inference with Centroid Learning and Table Lookup](https://mlsanity.com/paper/2302.03213/) (2023)
- [AdaptiveNet: Post-deployment Neural Architecture Adaptation for Diverse Edge Environments](https://mlsanity.com/paper/2303.07129/) (2023)
- [Variational Information Maximization for Feature Selection](https://mlsanity.com/paper/1606.02827/) (2016)
- [To Store or Not? Online Data Selection for Federated Learning with Limited Storage](https://mlsanity.com/paper/2209.00195/) (2022)
- [FedConv: A Learning-on-Model Paradigm for Heterogeneous Federated Clients](https://mlsanity.com/paper/2502.20639/) (2025)
- [Delta: A Cloud-assisted Data Enrichment Framework for On-Device Continual Learning](https://mlsanity.com/paper/2410.18378/) (2024)
- [MobiRNN: Efficient Recurrent Neural Network Execution on Mobile GPU](https://mlsanity.com/paper/1706.00878/) (2017)
- [Active Feature Selection for the Mutual Information Criterion](https://mlsanity.com/paper/2012.06979/) (2020)
- [Optimizing Storage Overhead of User Behavior Log for ML-embedded Mobile Apps](https://mlsanity.com/paper/2510.13405/) (2025)

## Abstract (arXiv)

Machine learning models are widely integrated into modern mobile apps to analyze user behaviors and deliver personalized services. Ensuring low-latency on-device model execution is critical for maintaining high-quality user experiences. While prior research has primarily focused on accelerating model inference with given input features, we identify an overlooked bottleneck in real-world on-device model execution pipelines: extracting input features from raw application logs. In this work, we explore a new direction of feature extraction optimization by analyzing and eliminating redundant extraction operations across different model features and consecutive model inferences. We then introduce AutoFeature, an automated feature extraction engine designed to accelerate on-device feature extraction process without compromising model inference accuracy. AutoFeature comprises three core designs: (1) graph abstraction to formulate the extraction workflows of different input features as one directed acyclic graph, (2) graph optimization to identify and fuse redundant operation nodes across different features within the graph; (3) efficient caching to minimize operations on overlapping raw data between consecutive model inferences. We implement a system prototype of AutoFeature and integrate it into five industrial mobile services spanning search, video and e-commerce domains. Online evaluations show that AutoFeature reduces end-to-end on-device model execution latency by 1.33x-3.93x during daytime and 1.43x-4.53x at night.