Mike Papadakis

Trajectory prediction models have become increasingly vital in various safety-critical domains, such as healthcare, maritime surveillance, and intelligent transportation systems. These applications utilize machine learning models to predict the operational state of a moving agent based on its motion trajectory. Despite their widespread deployment in these domains, testing such models remains a significant challenge due to the high cost of manual labeling, further exacerbated by the large scale and domain-specific characteristics of trajectory datasets. Test input prioritization has emerged as a promising solution to address the labeling cost issue, aiming to identify potentially misclassified inputs early to accelerate debugging and improve testing efficiency. The current state‑of‑the‑art test prioritization technique that can be applied to trajectory prediction tasks is MLPrior. However, MLPrior has the following limitations when applied to trajectory prediction scenarios: 1) MLPrior relies on rich input features, which are limited in trajectory data. Trajectory inputs typically contain only basic spatiotemporal coordinates (e.g., latitude, longitude, timestamps), offering insufficient attribute richness. 2) MLPrior inherits the limitations of learning‑based approaches. When the training data are imbalanced, their effectiveness can be significantly reduced. To overcome these limitations, we propose TrajectoryTest, a trajectory-specific test prioritization strategy that integrates trajectory-related information with MLPrior-derived representations to rank test inputs. Moreover, TrajectoryTest employs an adaptive strategy that switches between learning-based and uncertainty-based ranking depending on the model’s prediction error ratio. We conduct a comprehensive empirical study that confirms the limitations of MLPrior in trajectory prediction scenarios and demonstrates that TrajectoryTest outperforms all existing test input prioritization techniques, including the state-of-the-art approach MLPrior, multiple confidence-based methods, and the baseline random selection. The experimental results show that TrajectoryTest achieves improvements ranging from 7.03% to 9.56% over MLPrior (the state-of-the-art method) and confidence-based approaches on natural datasets, and from 6.71% to 9.94% on noisy datasets.

LLM-based mutation testing is a promising testing technology, but existing approaches typically rely on a fixed set of mutants as few-shot examples or none at all. This can result in generic low-quality mutants, missed context-specific mutation patterns, substantial numbers of redundant and non-compilable mutants, and limited semantic similarity to real bugs. To overcome these limitations, we introduce SMART (Semantic Mutation with Adaptive Retrieval and Tuning). SMART integrates retrieval-augmented generation (RAG) on a vectorized dataset of real-world bugs, focused code chunking, and supervised fine-tuning using mutants coupled with real-world bugs. We conducted an extensive empirical study of SMART using 1,991 real-world Java bugs from the Defects4J and ConDefects datasets, comparing SMART to the state-of-the-art LLM-based approaches, LLMut and LLMorpheus. The results reveal that SMART substantially improves mutation validity, effectiveness, and efficiency (even enabling 7B-scale models to match or even surpass large models like GPT-4o). We also demonstrate that SMART significantly improves downstream software engineering applications, including test case prioritization and fault localization. More specifically, SMART improves validity (weighted average generation rate) from 42.89% to 65.6%. It raises the non-duplicate rate from 87.38% to 95.62%, and the compilable rate from 88.85% to 90.21%. In terms of effectiveness, it achieves a real bug detection rate of 92.61% (vs. 57.86% for LLMut) and improves the average Ochiai coefficient from 25.61% to 38.44%. For fault localization, SMART ranks 64 more bugs as Top-1 under MUSE and 57 more under Metallaxis.

Modern web applications evolve continuously, often causing test breakages when test scripts fail to adapt to changes in the Document Object Model (DOM). Such breakages frequently arise from web locators, which identify the DOM elements that tests interact with. Since locators rely on specific elements and attributes to navigate the DOM, they are inherently vulnerable to changes. To address this, we propose EAGL, a technique for generating robust locators by leveraging the stability of DOM positions—resolving to either element or attribute nodes—predicted from historical DOM changes. EAGL synthesises locators by dynamically selecting stable and unique positions adequate to compose a locator for an element. We constructed a dataset of DOM position changes across 181 real-world websites and trained a stability prediction model that identifies persistent positions. Experimental results show that EAGL achieves strong stability prediction performance, with high classification metrics (precision 0.87, recall 0.85, F1 0.80, ROC-AUC 0.71) and robust ranking capability (MAP 0.83), reflecting its ability to prioritise stable DOM positions. Leveraging these models, EAGL generates robust and precise locators, outperforming Selenium and ROBULA+ while maintaining high efficiency (average locator generation time under 0.25 s). Cross-project evaluations confirm that EAGL generalises well across different websites, demonstrating its practical usefulness.

Web applications evolve rapidly, and UI changes cause automated end-to-end tests to fail. This paper reports a conceptual, independent replication of the COLOR locator repair approach. We reimplemented COLOR from its published description and deployed our implementation in an industrial Continuous Integration (CI) system, where tests are executed automatically on every codebase change. We analysed 4,471 test executions across 133 CI days spanning over seven months, categorised recurring failure modes, and found locator breakage and element visibility issues to be the dominant causes of instability. We analysed failure patterns introduced with evolution, collected element descriptors at runtime and deployed replication of the locator repair technique (COLOR) when locators fail. The repair mechanism suggested fixes for all captured broken locators, with the top recommendation correct in most cases, while exposing limitations caused by initially missing element data, dynamically generated locators, and sparse or volatile HTML attributes. Our results quantify the impact of attribute composition on repairability and highlight practical integration challenges for automated repair in daily CI workflows.

Distribution shift poses a significant challenge for deep learning (DL) models in source code analysis, where test data often follows different distributions from training data, leading to unexpected performance degradation and hindering the practical usage of code models. While our previous work CodeS introduced the first benchmark for studying distribution shift in source code analysis, it has limitations in covering more fine-grained types of real-world distribution shifts and lacks the study of the effectiveness of shift mitigation strategies. In this paper, we present CodeS+, an enhanced benchmark that addresses these limitations through two key contributions, (1) expanded shift types, we design more fine-grained distribution shift types, that is, shift introduced by different program element complexity (e.g., different node number of control flow graphs), and (2) investigate the usefulness of fine-tuning based shift mitigation techniques, such as Core-Set. Comprehensive experiments on different pre-trained code models demonstrated that code models significantly suffer from distribution shift, out-of-distribution detectors from other domains (e.g., computer vision) do not generalize to source code, and existing fine-tuning based shift mitigation techniques have limited benefits in enhancing the generalization ability of code models. Our findings highlight the need to pay more attention to OOD issues for code models.