Umar Farooq

M. H. Maqbool, Umar Farooq, Adib Mosharrof et al.

Recommender systems have become ubiquitous in our digital lives, from recommending products on e-commerce websites to suggesting movies and music on streaming platforms. Existing recommendation datasets, such as Amazon Product Reviews and MovieLens, greatly facilitated the research and development of recommender systems in their respective domains. While the number of mobile users and applications (aka apps) has increased exponentially over the past decade, research in mobile app recommender systems has been significantly constrained, primarily due to the lack of high-quality benchmark datasets, as opposed to recommendations for products, movies, and news. To facilitate research for app recommendation systems, we introduce a large-scale dataset, called MobileRec. We constructed MobileRec from users' activity on the Google play store. MobileRec contains 19.3 million user interactions (i.e., user reviews on apps) with over 10K unique apps across 48 categories. MobileRec records the sequential activity of a total of 0.7 million distinct users. Each of these users has interacted with no fewer than five distinct apps, which stands in contrast to previous datasets on mobile apps that recorded only a single interaction per user. Furthermore, MobileRec presents users' ratings as well as sentiments on installed apps, and each app contains rich metadata such as app name, category, description, and overall rating, among others. We demonstrate that MobileRec can serve as an excellent testbed for app recommendation through a comparative study of several state-of-the-art recommendation approaches. The quantitative results can act as a baseline for other researchers to compare their results against. The MobileRec dataset is available at https://huggingface.co/datasets/recmeapp/mobilerec.

Daniel Liu, Krishna Upadhyay, Vinaik Chhetri et al.

The rapid emergence of multi-agent AI systems (MAS), including LangChain, CrewAI, and AutoGen, has shaped how large language model (LLM) applications are developed and orchestrated. However, little is known about how these systems evolve and are maintained in practice. This paper presents the first large-scale empirical study of open-source MAS, analyzing over 42K unique commits and over 4.7K resolved issues across eight leading systems. Our analysis identifies three distinct development profiles: sustained, steady, and burst-driven. These profiles reflect substantial variation in ecosystem maturity. Perfective commits constitute 40.8% of all changes, suggesting that feature enhancement is prioritized over corrective maintenance (27.4%) and adaptive updates (24.3%). Data about issues shows that the most frequent concerns involve bugs (22%), infrastructure (14%), and agent coordination challenges (10%). Issue reporting also increased sharply across all frameworks starting in 2023. Median resolution times range from under one day to about two weeks, with distributions skewed toward fast responses but a minority of issues requiring extended attention. These results highlight both the momentum and the fragility of the current ecosystem, emphasizing the need for improved testing infrastructure, documentation quality, and maintenance practices to ensure long-term reliability and sustainability.

Vinaik Chhetri, A. B Siddique, Umar Farooq

Large language models (LLMs) are increasingly used in software development. However, while LLMs remain static after pretraining, programming languages and APIs continue to evolve, leading to the generation of deprecated or incompatible code that undermines reliability. Retraining LLMs from scratch to reflect such changes is computationally expensive, making model editing a promising lightweight alternative that updates only a small subset of parameters. Despite its potential, it remains unclear whether model editing yields genuine syntactic and semantic adaptations or merely superficial fixes. In this work, we present a systematic study of five state-of-the-art model editing methods: Constrained Fine-Tuning (FT), GRACE, MEMIT, PMET, and ROME. We apply these methods to three leading open-source code LLMs, CodeLlama, CodeQwen1.5, and DeepSeek-Coder, under controlled API deprecation scenarios. Our evaluation covers both instant and sequential editing settings, using three disjoint evaluation sets designed to assess reliability, generalization, and specificity. We measure model correctness at three levels: successful compilation, partial test case pass, and full test pass. Our findings show that instant edits consistently degrade model performance, with syntactic validity dropping by up to 86 percentage points and functional correctness declining by 45 points even in the best-performing setting. Sequential edits further amplify this degradation, and in some cases, model performance collapses entirely. Across all models, most passing generations relied on workarounds rather than correctly adopting the intended changes, while faulty adoptions that result in test failures or compilation errors were significantly more frequent. Correct adoptions, where the model correctly integrates the intended change, occurred in only about 6% of cases.

Moghis Fereidouni, Adib Mosharrof, Umar Farooq et al.

Mobile app stores produce a tremendous amount of data in the form of user reviews, which is a huge source of user requirements and sentiments; such reviews allow app developers to proactively address issues in their apps. However, only a small number of reviews capture common issues and sentiments which creates a need for automatically identifying prominent reviews. Unfortunately, most existing work in text ranking and popularity prediction focuses on social contexts where other signals are available, which renders such works ineffective in the context of app reviews. In this work, we propose a new framework, PPrior, that enables proactive prioritization of app issues through identifying prominent reviews (ones predicted to receive a large number of votes in a given time window). Predicting highly-voted reviews is challenging given that, unlike social posts, social network features of users are not available. Moreover, there is an issue of class imbalance, since a large number of user reviews receive little to no votes. PPrior employs a pre-trained T5 model and works in three phases. Phase one adapts the pre-trained T5 model to the user reviews data in a self-supervised fashion. In phase two, we leverage contrastive training to learn a generic and task-independent representation of user reviews. Phase three uses radius neighbors classifier t o m ake t he final predictions. This phase also uses FAISS index for scalability and efficient search. To conduct extensive experiments, we acquired a large dataset of over 2.1 million user reviews from Google Play. Our experimental results demonstrate the effectiveness of the proposed framework when compared against several state-of-the-art approaches. Moreover, the accuracy of PPrior in predicting prominent reviews is comparable to that of experienced app developers.

Muhammad Umair Haider, Umar Farooq, A. B. Siddique et al.

Language Models (LMs) have shown their application for tasks pertinent to code and several code~LMs have been proposed recently. The majority of the studies in this direction only focus on the improvements in performance of the LMs on different benchmarks, whereas LMs are considered black boxes. Besides this, a handful of works attempt to understand the role of attention layers in the code~LMs. Nonetheless, feed-forward layers remain under-explored which consist of two-thirds of a typical transformer model's parameters. In this work, we attempt to gain insights into the inner workings of code language models by examining the feed-forward layers. To conduct our investigations, we use two state-of-the-art code~LMs, Codegen-Mono and Ploycoder, and three widely used programming languages, Java, Go, and Python. We focus on examining the organization of stored concepts, the editability of these concepts, and the roles of different layers and input context size variations for output generation. Our empirical findings demonstrate that lower layers capture syntactic patterns while higher layers encode abstract concepts and semantics. We show concepts of interest can be edited within feed-forward layers without compromising code~LM performance. Additionally, we observe initial layers serve as ``thinking'' layers, while later layers are crucial for predicting subsequent code tokens. Furthermore, we discover earlier layers can accurately predict smaller contexts, but larger contexts need critical later layers' contributions. We anticipate these findings will facilitate better understanding, debugging, and testing of code~LMs.

Krishna Upadhyay, Moshood Fakorede, Umar Farooq

Quantum simulators are a foundational component of the quantum software ecosystem. They are widely used to develop and debug quantum programs, validate compiler transformations, and support empirical claims about correctness and performance. In the absence of large-scale quantum hardware, simulator outputs are often treated as ground truth for algorithm development and system evaluation. However, quantum simulators also introduce unique implementation challenges. They must faithfully emulate quantum behavior while executing on classical hardware, requiring complex representations of quantum state evolution, operator composition, and noise modeling. Yet, we still lack a large-scale and in-depth study of failures in quantum simulators. To bridge this gap, this work presents a comprehensive empirical study of bugs in widely used open-source quantum simulators. We analyze 394 confirmed bugs from 12 simulators and manually categorize them based on root causes, failure manifestations, affected components, and discovery mechanisms. Our study reveals several key findings. First, bug discovery is largely user-driven, with most crashes, exceptions, and resource-related failures not detected by automated testing and identified after deployment. Second, logical correctness failures are widespread and often silent, producing plausible but incorrect outputs without triggering crashes or explicit error signals. Third, many critical failures originate in classical simulator infrastructure, such as memory management, indexing, configuration, and dependency compatibility, rather than in core quantum execution logic. These findings provide new insights into the reliability challenges of quantum simulators and highlight opportunities to improve testing and validation practices in the quantum software ecosystem.

Moshood A. Fakorede, Krishna Upadhyay, A. B. Siddique et al.

Large language models (LLMs) have shown strong performance on automated software engineering tasks, yet existing benchmarks focus primarily on general-purpose libraries or web applications, leaving mobile application development largely unexplored despite its strict platform constraints, framework-driven lifecycles, and complex platform API interactions. We introduce MobileDev-Bench, a benchmark comprising 384 real-world issue-resolution tasks collected from 18 production mobile applications spanning Android Native (Java/Kotlin), React Native (TypeScript), and Flutter (Dart). Each task pairs an authentic developer-reported issue with executable test patches, enabling fully automated validation of model-generated fixes within mobile build environments. The benchmark exhibits substantial patch complexity: fixes modify 12.5 files and 324.9 lines on average, and 35.7% of instances require coordinated changes across multiple artifact types, such as source and manifest files. Evaluation of four state-of-the-art code-capable LLMs, GPT- 5.2, Claude Sonnet 4.5, Gemini Flash 2.5, and Qwen3-Coder, yields low end-to-end resolution rates of 3.39%-5.21%, revealing significant performance gaps compared to prior benchmarks. Further analysis reveals systematic failure modes, with fault localization across multi-file and multi-artifact changes emerging as the primary bottleneck.

M. H. Maqbool, Moghis Fereidouni, Umar Farooq et al.

The mobile app market has expanded exponentially, offering millions of apps with diverse functionalities, yet research in mobile app recommendation remains limited. Traditional sequential recommender systems utilize the order of items in users' historical interactions to predict the next item for the users. Position embeddings, well-established in transformer-based architectures for natural language processing tasks, effectively distinguish token positions in sequences. In sequential recommendation systems, position embeddings can capture the order of items in a user's historical interaction sequence. Nevertheless, this ordering does not consider the time elapsed between two interactions of the same user (e.g., 1 day, 1 week, 1 month), referred to as "user rhythm". In mobile app recommendation datasets, the time between consecutive user interactions is notably longer compared to other domains like movies, posing significant challenges for sequential recommender systems. To address this phenomenon in the mobile app domain, we introduce INTERPOS, an Interaction Rhythm Guided Positional Morphing strategy for autoregressive mobile app recommender systems. INTERPOS incorporates rhythm-guided position embeddings, providing a more comprehensive representation that considers both the sequential order of interactions and the temporal gaps between them. This approach enables a deep understanding of users' rhythms at a fine-grained level, capturing the intricacies of their interaction patterns over time. We propose three strategies to incorporate the morphed positional embeddings in two transformer-based sequential recommendation system architectures. Our extensive evaluations show that INTERPOS outperforms state-of-the-art models using 7 mobile app recommendation datasets on NDCG@K and HIT@K metrics. The source code of INTERPOS is available at https://github.com/dlgrad/INTERPOS.

Maëlick Claes, Mika Mäntylä, Miikka Kuutila et al.

Open source development contains contributions from both hired and volunteer software developers. Identification of this status is important when we consider the transferability of research results to the closed source software industry, as they include no volunteer developers. While many studies have taken the employment status of developers into account, this information is often gathered manually due to the lack of accurate automatic methods. In this paper, we present an initial step towards predicting paid and unpaid open source development using machine learning and compare our results with automatic techniques used in prior work. By relying on code source repository meta-data from Mozilla, and manually collected employment status, we built a dataset of the most active developers, both volunteer and hired by Mozilla. We define a set of metrics based on developers' usual commit time pattern and use different classification methods (logistic regression, classification tree, and random forest). The results show that our proposed method identify paid and unpaid commits with an AUC of 0.75 using random forest, which is higher than the AUC of 0.64 obtained with the best of the previously used automatic methods.

Umar Farooq, Jean-Yves Guillemaut, Adrian Hilton et al.

The field of Novel View Synthesis has been revolutionized by 3D Gaussian Splatting (3DGS), which enables high-quality scene reconstruction that can be rendered in real-time. 3DGS-based techniques typically suffer from high GPU memory and disk storage requirements which limits their practical application on consumer-grade devices. We propose Opti3DGS, a novel frequency-modulated coarse-to-fine optimization framework that aims to minimize the number of Gaussian primitives used to represent a scene, thus reducing memory and storage demands. Opti3DGS leverages image frequency modulation, initially enforcing a coarse scene representation and progressively refining it by modulating frequency details in the training images. On the baseline 3DGS, we demonstrate an average reduction of 62% in Gaussians, a 40% reduction in the training GPU memory requirements and a 20% reduction in optimization time without sacrificing the visual quality. Furthermore, we show that our method integrates seamlessly with many 3DGS-based techniques, consistently reducing the number of Gaussian primitives while maintaining, and often improving, visual quality. Additionally, Opti3DGS inherently produces a level-of-detail scene representation at no extra cost, a natural byproduct of the optimization pipeline. Results and code will be made publicly available.

Vinaik Chhetri, Yousaf Reza, Moghis Fereidouni et al.

Modern recommendation systems typically follow two complementary paradigms: collaborative filtering, which models long-term user preferences from historical interactions, and conversational recommendation systems (CRS), which interact with users in natural language to uncover immediate needs. Each captures a different dimension of user intent. While CRS models lack collaborative signals, leading to generic or poorly personalized suggestions, traditional recommenders lack mechanisms to interactively elicit immediate needs. Unifying these paradigms promises richer personalization but remains challenging due to the lack of large-scale conversational datasets grounded in real user behavior. We present ConvRecStudio, a framework that uses large language models (LLMs) to simulate realistic, multi-turn dialogs grounded in timestamped user-item interactions and reviews. ConvRecStudio follows a three-stage pipeline: (1) Temporal Profiling, which constructs user profiles and community-level item sentiment trajectories over fine-grained aspects; (2) Semantic Dialog Planning, which generates a structured plan using a DAG of flexible super-nodes; and (3) Multi-Turn Simulation, which instantiates the plan using paired LLM agents for the user and system, constrained by executional and behavioral fidelity checks. We apply ConvRecStudio to three domains -- MobileRec, Yelp, and Amazon Electronics -- producing over 12K multi-turn dialogs per dataset. Human and automatic evaluations confirm the naturalness, coherence, and behavioral grounding of the generated conversations. To demonstrate utility, we build a cross-attention transformer model that jointly encodes user history and dialog context, achieving gains in Hit@K and NDCG@K over baselines using either signal alone or naive fusion. Notably, our model achieves a 10.9% improvement in Hit@1 on Yelp over the strongest baseline.

Vinaik Chhetri, Krishna Upadhyay, A. B. Siddique et al.

Artificial Intelligence (AI)-powered features have rapidly proliferated across mobile apps in various domains, including productivity, education, entertainment, and creativity. However, how users perceive, evaluate, and critique these AI features remains largely unexplored, primarily due to the overwhelming volume of user feedback. In this work, we present the first comprehensive, large-scale study of user feedback on AI-powered mobile apps, leveraging a curated dataset of 292 AI-driven apps across 14 categories with 894K AI-specific reviews from Google Play. We develop and validate a multi-stage analysis pipeline that begins with a human-labeled benchmark and systematically evaluates large language models (LLMs) and prompting strategies. Each stage, including review classification, aspect-sentiment extraction, and clustering, is validated for accuracy and consistency. Our pipeline enables scalable, high-precision analysis of user feedback, extracting over one million aspect-sentiment pairs clustered into 18 positive and 15 negative user topics. Our analysis reveals that users consistently focus on a narrow set of themes: positive comments emphasize productivity, reliability, and personalized assistance, while negative feedback highlights technical failures (e.g., scanning and recognition), pricing concerns, and limitations in language support. Our pipeline surfaces both satisfaction with one feature and frustration with another within the same review. These fine-grained, co-occurring sentiments are often missed by traditional approaches that treat positive and negative feedback in isolation or rely on coarse-grained analysis. To this end, our approach provides a more faithful reflection of the real-world user experiences with AI-powered apps. Category-aware analysis further uncovers both universal drivers of satisfaction and domain-specific frustrations.

Umar Farooq, A. B. Siddique, Fuad Jamour et al.

Responding to user reviews promptly and satisfactorily improves application ratings, which is key to application popularity and success. The proliferation of such reviews makes it virtually impossible for developers to keep up with responding manually. To address this challenge, recent work has shown the possibility of automatic response generation. However, because the training review-response pairs are aggregated from many different apps, it remains challenging for such models to generate app-specific responses, which, on the other hand, are often desirable as apps have different features and concerns. Solving the challenge by simply building a model per app (i.e., training with review-response pairs of a single app) may be insufficient because individual apps have limited review-response pairs, and such pairs typically lack the relevant information needed to respond to a new review. To enable app-specific response generation, this work proposes AARSynth: an app-aware response synthesis system. The key idea behind AARSynth is to augment the seq2seq model with information specific to a given app. Given a new user review, it first retrieves the top-K most relevant app reviews and the most relevant snippet from the app description. The retrieved information and the new user review are then fed into a fused machine learning model that integrates the seq2seq model with a machine reading comprehension model. The latter helps digest the retrieved reviews and app description. Finally, the fused model generates a response that is customized to the given app. We evaluated AARSynth using a large corpus of reviews and responses from Google Play. The results show that AARSynth outperforms the state-of-the-art system by 22.2% on BLEU-4 score. Furthermore, our human study shows that AARSynth produces a statistically significant improvement in response quality compared to the state-of-the-art system.

Miikka Kuutila, Mika Mäntylä, Umar Farooq et al.

Large project overruns and overtime work have been reported in the software industry, resulting in additional expense for companies and personal issues for developers. The present work aims to provide an overview of studies related to time pressure in software engineering; specifically, existing definitions, possible causes, and metrics relevant to time pressure were collected, and a mapping of the studies to software processes and approaches was performed. Moreover, we synthesize results of existing quantitative studies on the effects of time pressure on software development, and offer practical takeaways for practitioners and researchers, based on empirical evidence. Our search strategy examined 5,414 sources, found through repository searches and snowballing. Applying inclusion and exclusion criteria resulted in the selection of 102 papers, which made relevant contributions related to time pressure in software engineering. The majority of high quality studies report increased productivity and decreased quality under time pressure. Frequent categories of studies focus on quality assurance, cost estimation, and process simulation. It appears that time pressure is usually caused by errors in cost estimation. The effect of time pressure is most often identified during software quality assurance. The majority of empirical studies report increased productivity under time pressure, while the most cost estimation and process simulation models assume that compressing the schedule increases the total needed hours. We also find evidence of the mediating effect of knowledge on the effects of time pressure, and that tight deadlines impact tasks with an algorithmic nature more severely. Future research should better contextualize quantitative studies to account for the existing conflicting results and to provide an understanding of situations when time pressure is either beneficial or harmful.

Maëlick Claes, Mika Mäntylä, Umar Farooq

Background: Using sentiment analysis to study software developers' behavior comes with challenges such as the presence of a large amount of technical discussion unlikely to express any positive or negative sentiment. However, emoticons provide information about developer sentiments that can easily be extracted from software repositories. Aim: We investigate how software developers use emoticons differently in issue trackers in order to better understand the differences between developers and determine to which extent emoticons can be used as in place of sentiment analysis. Method: We extract emoticons from 1.3M comments from Apache's issue tracker and 4.5M from Mozilla's issue tracker using regular expressions built from a list of emoticons used by SentiStrength and Wikipedia. We check for statistical differences using Mann-Whitney U tests and determine the effect size with Cliff's delta. Results: Overall Mozilla developers rely more on emoticons than Apache developers. While the overall ratio of comments with emoticons is of 2% and 3.6% for Apache and Mozilla, some individual developers can have a ratio above 20%. Looking specifically at Mozilla developers, we find that western developers use significantly more emoticons (with large size effect) than eastern developers. While the majority of emoticons are used to express joy, we find that Mozilla developers use emoticons more frequently to express sadness and surprise than Apache developers. Finally, we find that developers use overall more emoticons during weekends than during weekdays, with the share of sad and surprised emoticons increasing during weekends. Conclusions: While emoticons are primarily used to express joy, the more occasional use of sad and surprised emoticons can potentially be utilized to detect frustration in place of sentiment analysis among developers using emoticons frequently enough.

Mika Mäntylä, Maëlick Claes, Umar Farooq

Background: Unstructured and textual data is increasing rapidly and Latent Dirichlet Allocation (LDA) topic modeling is a popular data analysis methods for it. Past work suggests that instability of LDA topics may lead to systematic errors. Aim: We propose a method that relies on replicated LDA runs, clustering, and providing a stability metric for the topics. Method: We generate k LDA topics and replicate this process n times resulting in n*k topics. Then we use K-medioids to cluster the n*k topics to k clusters. The k clusters now represent the original LDA topics and we present them like normal LDA topics showing the ten most probable words. For the clusters, we try multiple stability metrics, out of which we recommend Rank-Biased Overlap, showing the stability of the topics inside the clusters. Results: We provide an initial validation where our method is used for 270,000 Mozilla Firefox commit messages with k=20 and n=20. We show how our topic stability metrics are related to the contents of the topics. Conclusions: Advances in text mining enable us to analyze large masses of text in software engineering but non-deterministic algorithms, such as LDA, may lead to unreplicable conclusions. Our approach makes LDA stability transparent and is also complementary rather than alternative to many prior works that focus on LDA parameter tuning.