With the rapid growth in the number of Large Language Models (LLMs), there has been a recent interest in LLM routing, or directing queries to the cheapest LLM that can deliver a suitable response. We conduct a minimax analysis of the routing problem, providing a lower bound and finding that a simple router that predicts both cost and accuracy for each question can be minimax optimal. Inspired by this, we introduce CARROT, a Cost AwaRe Rate Optimal rouTer that selects a model based on estimates of the models' cost and performance. Alongside CARROT, we also introduce the Smart Price-aware ROUTing (SPROUT) dataset to facilitate routing on a wide spectrum of queries with the latest state-of-the-art LLMs. Using SPROUT and prior benchmarks such as Routerbench and open-LLM-leaderboard-v2 we empirically validate CARROT's performance against several alternative routers.

Large Language Models (LLMs) have gained attention for addressing coding problems, but their effectiveness in fixing code maintainability remains unclear. This study evaluates LLMs capability to resolve 127 maintainability issues from 10 GitHub repositories. We use zero-shot prompting for Copilot Chat and Llama 3.1, and few-shot prompting with Llama only. The LLM-generated solutions are assessed for compilation errors, test failures, and new maintainability problems. Llama with few-shot prompting successfully fixed 44.9% of the methods, while Copilot Chat and Llama zero-shot fixed 32.29% and 30%, respectively. However, most solutions introduced errors or new maintainability issues. We also conducted a human study with 45 participants to evaluate the readability of 51 LLM-generated solutions. The human study showed that 68.63% of participants observed improved readability. Overall, while LLMs show potential for fixing maintainability issues, their introduction of errors highlights their current limitations.

Recent advancements in super-resolution for License Plate Recognition (LPR) have sought to address challenges posed by low-resolution (LR) and degraded images in surveillance, traffic monitoring, and forensic applications. However, existing studies have relied on private datasets and simplistic degradation models. To address this gap, we introduce UFPR-SR-Plates, a novel dataset containing 10,000 tracks with 100,000 paired low and high-resolution license plate images captured under real-world conditions. We establish a benchmark using multiple sequential LR and high-resolution (HR) images per vehicle -- five of each -- and two state-of-the-art models for super-resolution of license plates. We also investigate three fusion strategies to evaluate how combining predictions from a leading Optical Character Recognition (OCR) model for multiple super-resolved license plates enhances overall performance. Our findings demonstrate that super-resolution significantly boosts LPR performance, with further improvements observed when applying majority vote-based fusion techniques. Specifically, the Layout-Aware and Character-Driven Network (LCDNet) model combined with the Majority Vote by Character Position (MVCP) strategy led to the highest recognition rates, increasing from 1.7% with low-resolution images to 31.1% with super-resolution, and up to 44.7% when combining OCR outputs from five super-resolved images. These findings underscore the critical role of super-resolution and temporal information in enhancing LPR accuracy under real-world, adverse conditions. The proposed dataset is publicly available to support further research and can be accessed at: this https URL

Despite significant advancements in License Plate Recognition (LPR) through deep learning, most improvements rely on high-resolution images with clear characters. This scenario does not reflect real-world conditions where traffic surveillance often captures low-resolution and blurry images. Under these conditions, characters tend to blend with the background or neighboring characters, making accurate LPR challenging. To address this issue, we introduce a novel loss function, Layout and Character Oriented Focal Loss (LCOFL), which considers factors such as resolution, texture, and structural details, as well as the performance of the LPR task itself. We enhance character feature learning using deformable convolutions and shared weights in an attention module and employ a GAN-based training approach with an Optical Character Recognition (OCR) model as the discriminator to guide the super-resolution process. Our experimental results show significant improvements in character reconstruction quality, outperforming two state-of-the-art methods in both quantitative and qualitative measures. Our code is publicly available at this https URL

Software development increasingly depends on libraries and frameworks to increase productivity and reduce time-to-market. Despite this fact, we still lack techniques to assess developers expertise in widely popular libraries and frameworks. In this paper, we evaluate the performance of unsupervised (based on clustering) and supervised machine learning classifiers (Random Forest and SVM) to identify experts in three popular JavaScript libraries: facebook/react, mongodb/node-mongodb, and socketio/socket.io. First, we collect 13 features about developers activity on GitHub projects, including commits on source code files that depend on these libraries. We also build a ground truth including the expertise of 575 developers on the studied libraries, as self-reported by them in a survey. Based on our findings, we document the challenges of using machine learning classifiers to predict expertise in software libraries, using features extracted from GitHub. Then, we propose a method to identify library experts based on clustering feature data from GitHub; by triangulating the results of this method with information available on Linkedin profiles, we show that it is able to recommend dozens of GitHub users with evidences of being experts in the studied JavaScript libraries. We also provide a public dataset with the expertise of 575 developers on the studied libraries.

Despite being introduced only a few years ago, Large Language Models (LLMs) are already widely used by developers for code generation. However, their application in automating other Software Engineering activities remains largely unexplored. Thus, in this paper, we report the first results of a study in which we are exploring the use of ChatGPT to support API migration tasks, an important problem that demands manual effort and attention from developers. Specifically, in the paper, we share our initial results involving the use of ChatGPT to migrate a client application to use a newer version of SQLAlchemy, an ORM (Object Relational Mapping) library widely used in Python. We evaluate the use of three types of prompts (Zero-Shot, One-Shot, and Chain Of Thoughts) and show that the best results are achieved by the One-Shot prompt, followed by the Chain Of Thoughts. Particularly, with the One-Shot prompt we were able to successfully migrate all columns of our target application and upgrade its code to use new functionalities enabled by SQLAlchemy's latest version, such as Python's asyncio and typing modules, while preserving the original code behavior.

This paper introduces a novel Graph Neural Network (GNN) architecture for time series classification, based on visibility graph representations. Traditional time series classification methods often struggle with high computational complexity and inadequate capture of spatio-temporal dynamics. By representing time series as visibility graphs, it is possible to encode both spatial and temporal dependencies inherent to time series data, while being computationally efficient. Our architecture is fully modular, enabling flexible experimentation with different models and representations. We employ directed visibility graphs encoded with in-degree and PageRank features to improve the representation of time series, ensuring efficient computation while enhancing the model's ability to capture long-range dependencies in the data. We show the robustness and generalization capability of the proposed architecture across a diverse set of classification tasks and against a traditional model. Our work represents a significant advancement in the application of GNNs for time series analysis, offering a powerful and flexible framework for future research and practical implementations.

Robots must operate safely when deployed in novel and human-centered environments, like homes. Current safe control approaches typically assume that the safety constraints are known a priori, and thus, the robot can pre-compute a corresponding safety controller. While this may make sense for some safety constraints (e.g., avoiding collision with walls by analyzing a floor plan), other constraints are more complex (e.g., spills), inherently personal, context-dependent, and can only be identified at deployment time when the robot is interacting in a specific environment and with a specific person (e.g., fragile objects, expensive rugs). Here, language provides a flexible mechanism to communicate these evolving safety constraints to the robot. In this work, we use vision language models (VLMs) to interpret language feedback and the robot's image observations to continuously update the robot's representation of safety constraints. With these inferred constraints, we update a Hamilton-Jacobi reachability safety controller online via efficient warm-starting techniques. Through simulation and hardware experiments, we demonstrate the robot's ability to infer and respect language-based safety constraints with the proposed approach.

Automated news credibility and fact-checking at scale require accurately predicting news factuality and media bias. This paper introduces a large sentence-level dataset, titled "FactNews", composed of 6,191 sentences expertly annotated according to factuality and media bias definitions proposed by AllSides. We use FactNews to assess the overall reliability of news sources, by formulating two text classification problems for predicting sentence-level factuality of news reporting and bias of media outlets. Our experiments demonstrate that biased sentences present a higher number of words compared to factual sentences, besides having a predominance of emotions. Hence, the fine-grained analysis of subjectivity and impartiality of news articles provided promising results for predicting the reliability of media outlets. Finally, due to the severity of fake news and political polarization in Brazil, and the lack of research for Portuguese, both dataset and baseline were proposed for Brazilian Portuguese.

Code review is a key development practice that contributes to improve software quality and to foster knowledge sharing among developers. However, code review usually takes time and demands detailed and time-consuming analysis of textual diffs. Particularly, detecting refactorings during code reviews is not a trivial task, since they are not explicitly represented in diffs. For example, a Move Function refactoring is represented by deleted (-) and added lines (+) of code which can be located in different and distant source code files. To tackle this problem, we introduce RAID, a refactoring-aware and intelligent diff tool. Besides proposing an architecture for RAID, we implemented a Chrome browser plug-in that supports our solution. Then, we conducted a field experiment with eight professional developers who used RAID for three months. We concluded that RAID can reduce the cognitive effort required for detecting and reviewing refactorings in textual diff. Besides documenting refactorings in diffs, RAID reduces the number of lines required for reviewing such operations. For example, the median number of lines to be reviewed decreases from 14.5 to 2 lines in the case of move refactorings and from 113 to 55 lines in the case of extractions.

In the last few years thousands of scientific papers have investigated sentiment analysis, several startups that measure opinions on real data have emerged and a number of innovative products related to this theme have been developed. There are multiple methods for measuring sentiments, including lexical-based and supervised machine learning methods. Despite the vast interest on the theme and wide popularity of some methods, it is unclear which one is better for identifying the polarity (i.e., positive or negative) of a message. Accordingly, there is a strong need to conduct a thorough apple-to-apple comparison of sentiment analysis methods, \textit{as they are used in practice}, across multiple datasets originated from different data sources. Such a comparison is key for understanding the potential limitations, advantages, and disadvantages of popular methods. This article aims at filling this gap by presenting a benchmark comparison of twenty-four popular sentiment analysis methods (which we call the state-of-the-practice methods). Our evaluation is based on a benchmark of eighteen labeled datasets, covering messages posted on social networks, movie and product reviews, as well as opinions and comments in news articles. Our results highlight the extent to which the prediction performance of these methods varies considerably across datasets. Aiming at boosting the development of this research area, we open the methods' codes and datasets used in this article, deploying them in a benchmark system, which provides an open API for accessing and comparing sentence-level sentiment analysis methods.

Neoadjuvant chemotherapy (NAC) response prediction for triple negative breast cancer (TNBC) patients is a challenging task clinically as it requires understanding complex histology interactions within the tumor microenvironment (TME). Digital whole slide images (WSIs) capture detailed tissue information, but their giga-pixel size necessitates computational methods based on multiple instance learning, which typically analyze small, isolated image tiles without the spatial context of the TME. To address this limitation and incorporate TME spatial histology interactions in predicting NAC response for TNBC patients, we developed a histology context-aware transformer graph convolution network (NACNet). Our deep learning method identifies the histopathological labels on individual image tiles from WSIs, constructs a spatial TME graph, and represents each node with features derived from tissue texture and social network analysis. It predicts NAC response using a transformer graph convolution network model enhanced with graph isomorphism network layers. We evaluate our method with WSIs of a cohort of TNBC patient (N=105) and compared its performance with multiple state-of-the-art machine learning and deep learning models, including both graph and non-graph approaches. Our NACNet achieves 90.0% accuracy, 96.0% sensitivity, 88.0% specificity, and an AUC of 0.82, through eight-fold cross-validation, outperforming baseline models. These comprehensive experimental results suggest that NACNet holds strong potential for stratifying TNBC patients by NAC response, thereby helping to prevent overtreatment, improve patient quality of life, reduce treatment cost, and enhance clinical outcomes, marking an important advancement toward personalized breast cancer treatment.

In this paper, we tackle Automatic Meter Reading (AMR) by leveraging the high capability of Convolutional Neural Networks (CNNs). We design a two-stage approach that employs the Fast-YOLO object detector for counter detection and evaluates three different CNN-based approaches for counter recognition. In the AMR literature, most datasets are not available to the research community since the images belong to a service company. In this sense, we introduce a new public dataset, called UFPR-AMR dataset, with 2,000 fully and manually annotated images. This dataset is, to the best of our knowledge, three times larger than the largest public dataset found in the literature and contains a well-defined evaluation protocol to assist the development and evaluation of AMR methods. Furthermore, we propose the use of a data augmentation technique to generate a balanced training set with many more examples to train the CNN models for counter recognition. In the proposed dataset, impressive results were obtained and a detailed speed/accuracy trade-off evaluation of each model was performed. In a public dataset, state-of-the-art results were achieved using less than 200 images for training.

Next-generation touristic services will rely on the advanced mobile networks' high bandwidth and low latency and the Multi-access Edge Computing (MEC) paradigm to provide fully immersive mobile experiences. As an integral part of travel planning systems, recommendation algorithms devise personalized tour itineraries for individual users considering the popularity of a city's Points of Interest (POIs) as well as the tourist preferences and constraints. However, in the context of next-generation touristic services, recommendation algorithms should also consider the applications (e.g., social network, mobile video streaming, mobile augmented reality) the tourist will consume in the POIs and the quality in which the MEC infrastructure will deliver such applications. In this paper, we address the joint problem of recommending personalized tour itineraries for tourists and efficiently allocating MEC resources for advanced touristic applications. We formulate an optimization problem that maximizes the itinerary of individual tourists while optimizing the resource allocation at the network edge. We then propose an exact algorithm that quickly solves the problem optimally, considering instances of realistic size. Using a real-world location-based photo-sharing database, we conduct and present an exploratory analysis to understand preferences and users' visiting patterns. Using this understanding, we propose a methodology to identify user interest in applications. Finally, we evaluate our algorithm using this dataset. Results show that our algorithm outperforms a modified version of a state-of-the-art solution for personalized tour itinerary recommendation, demonstrating gains up to 11% for resource allocation efficiency and 40% for user experience. In addition, our algorithm performs similarly to the modified state-of-the-art solution regarding traditional itinerary recommendation metrics.

This is the first work to investigate the effectiveness of BERT-based contextual embeddings in active learning (AL) tasks on cold-start scenarios, where traditional fine-tuning is infeasible due to the absence of labeled data. Our primary contribution is the proposal of a more robust fine-tuning pipeline - DoTCAL - that diminishes the reliance on labeled data in AL using two steps: (1) fully leveraging unlabeled data through domain adaptation of the embeddings via masked language modeling and (2) further adjusting model weights using labeled data selected by AL. Our evaluation contrasts BERT-based embeddings with other prevalent text representation paradigms, including Bag of Words (BoW), Latent Semantic Indexing (LSI), and FastText, at two critical stages of the AL process: instance selection and classification. Experiments conducted on eight ATC benchmarks with varying AL budgets (number of labeled instances) and number of instances (about 5,000 to 300,000) demonstrate DoTCAL's superior effectiveness, achieving up to a 33% improvement in Macro-F1 while reducing labeling efforts by half compared to the traditional one-step method. We also found that in several tasks, BoW and LSI (due to information aggregation) produce results superior (up to 59% ) to BERT, especially in low-budget scenarios and hard-to-classify tasks, which is quite surprising.

The accuracy of a classifier, when performing Pattern recognition, is mostly tied to the quality and representativeness of the input feature vector. Feature Selection is a process that allows for representing information properly and may increase the accuracy of a classifier. This process is responsible for finding the best possible features, thus allowing us to identify to which class a pattern belongs. Feature selection methods can be categorized as Filters, Wrappers, and Embed. This paper presents a survey on some Filters and Wrapper methods for handcrafted feature selection. Some discussions, with regard to the data structure, processing time, and ability to well represent a feature vector, are also provided in order to explicitly show how appropriate some methods are in order to perform feature selection. Therefore, the presented feature selection methods can be accurate and efficient if applied considering their positives and negatives, finding which one fits best the problem's domain may be the hardest task.

This work uses vector field inequalities (VFI) to prevent robot self-collisions and collisions with the workspace. Differently from previous approaches, the method is suitable for both velocity and torque-actuated robots. We propose a new distance function and its corresponding Jacobian in order to generate a VFI to limit the angle between two Pl\"ucker lines. This new VFI is used to prevent both undesired end-effector orientations and violation of joints limits. The proposed method is evaluated in a realistic simulation and on a real humanoid robot, showing that all constraints are respected while the robot performs a manipulation task.