一句话：提出 GraphDroid，用聚类记忆、异步意图生成与混合意图满足提升 LLM 移动应用 GUI 测试覆盖率并大幅降低成本。
移动应用 GUI 自动化测试难以覆盖需要多步操作序列的复杂功能，传统工具缺乏语义理解，而现有 LLM 测试工具存在三个问题：丢失历史上下文导致难以识别未覆盖功能、同步生成意图阻塞探索、以及每步都调用 LLM 带来高成本与高延迟。本文提出 GraphDroid，一个意图驱动的 GUI 测试框架：通过基于聚类的记忆机制从历史访问状态中识别未覆盖功能；采用异步意图生成范式消除延迟瓶颈；并设计混合意图满足策略，将复杂意图交给 LLM，而把简单意图下放给轻量启发式算法。作者在 41 个真实 Android 应用上与六个先进基线对比，GraphDroid 全面胜出，代码覆盖率最高提升 36.4%，成本不足最佳纯 LLM 基线的八分之一；同时发现 19 个缺陷并在 Themis 基准中检出 52 个崩溃中的 13 个，均超过所有基线，其中 7 个为未知缺陷，已有 4 个被确认修复。
Automated GUI testing is a widely adopted technique for ensuring mobile application quality by simulating user interactions to exercise functionalities. Despite the research breakthroughs in the past decades, covering complex functionalities that require multi-step action sequences still remains challenging. Traditional tools lack semantic understanding capability and can rarely synthesize such action sequences. Recent LLM-based tools can generate test intents describing target functionalities and leverage the LLM to fulfill the intents, but suffer from three key limitations: 1) loss of historical context for identifying uncovered functionalities, 2) synchronous intent generation that blocks exploration, and 3) per-step LLM-driven fulfillment incurring high cost and latency. To address these limitations, we propose GraphDroid, an intent-driven GUI testing framework that integrates a cluster-based memory mechanism to effectively identify uncovered functionalities from historically visited states for comprehensive application testing. For improving testing efficiency, GraphDroid adopts an asynchronous intent generation paradigm that eliminates the latency bottleneck and a hybrid intent fulfillment strategy that reserves the LLM for fulfilling complex intents while delegating simple intents to a lightweight heuristic algorithm. We evaluate GraphDroid on 41 real-world Android apps against six state-of-the-art baselines. Results show that GraphDroid outperforms all baselines, achieving up to 36.4% higher code coverage while incurring less than one eighth of the cost of the best pure LLM-based baseline. GraphDroid also exposes 19 bugs in the 41 apps and detects 13 of 52 crashes in the Themis bug benchmark, surpassing all the six baselines. Seven of the 19 bugs were previously unknown and we reported them to the developers. So far, four bugs have been confirmed and fixed.
@inproceedings{paperbot12468,
title = {GraphDroid: Asynchronous LLM-Based Mobile App GUI Testing via History-Aware Exploration and Hybrid Intent Fulfillment},
author = {Xiaolei Li and Jialun Cao and Zhijian Hou and Yuzhi Zhao and Yepang Liu and Shing-Chi Cheung},
booktitle = {ASE},
year = {2026},
doi = {10.1145/3832783.3837449}
}