Lero

Field-based contact sports, including American football and rugby are among those with the highest sport-related concussion (SRC) rates. Regarding SRC treatment, gradual return-to-play (RTP) protocols are commonly used, however variability in RTP protocols between sporting bodies and scholastic institutions makes accurately interpreting published SRC severity data difficult. Compare SRC recovery times, RTP protocols, and mandated sit-out period duration among adult amateur contact field sports. MEDLINE, CINAHL Complete, SPORTDiscus, APA PsycInfo, and APA PsycArticles were searched on May 13th, 2024. Eligibility criteria included (1) published in English in a peer-reviewed journal from 2015 to 2024, (2) adult population, (3) amateur, field-based contact sport athlete population, (4) diagnosed and/or suspected SRC, and (5) recovery time (days/weeks) data available. 19/6085 (< 1%) studies comprising 58 datasets met all inclusion criteria. The studies comprised American football (10/19, 52.6%), field-hockey (5/19, 26.3%), lacrosse (9/19, 47.4%), ladies gaelic football (1/19, 5.3%), rugby (8/19, 42.1%), soccer (10/19, 52.6%), and sprint football (1/19, 5.2%). Mandated sit-out periods ranged from 0 to 21 days across sports, and RTP protocol durations ranged from 5 to 21 days. The most common RTP protocol duration was 7 days (n = 31/58), and mandated sit-out period was no same-day return (n = 25/58). Field hockey had the longest reported SRC recovery duration (34 days), and rugby had the longest mandated sit-out period (21 days). Recovery time following concussion in contact sports should be interpreted in light of sport type, mandated sit-out policies, and RTP protocols. Empirical evidence should drive decisions in RTP protocols and sit-out periods, which are currently inconsistent between sports.

Most log-based anomaly detectors assume logs are stable, though in reality they are often unstable due to software or environmental changes. Anomaly detection on unstable logs (ULAD) is therefore a more realistic, yet under-investigated challenge. Current approaches predominantly employ machine learning (ML) models, which often require extensive labeled data for training. To mitigate data insufficiency, we propose FlexLog , a novel hybrid approach for ULAD that combines ML models — decision tree, k-nearest neighbors, and a feedforward neural network — with a Large Language Model (Mistral) through ensemble learning. FlexLog also incorporates a cache and retrieval-augmented generation (RAG) to further enhance efficiency and effectiveness. To evaluate FlexLog , we configured four datasets for ULAD, namely ADFA-U, LOGEVOL-U, SynHDFS-U, and SYNEVOL-U. FlexLog outperforms all baselines by at least 1.2 percentage points (pp) in F1 score while using much less labeled data (62.87 pp reduction). When trained on the same amount of data as the baselines, FlexLog achieves up to a 13 pp increase in F1 score on ADFA-U across varying training dataset sizes. Additionally, FlexLog maintains inference time under one second per log sequence, making it suitable for most applications, except latency-sensitive systems. Further analysis reveals the positive impact of FlexLog ’s key components: cache, RAG, and ensemble learning.

Large language models (LLMs) have enabled new tools in requirements engineering (RE), often in the form of intelligent agents or virtual assistants. These tools can transform how software engineers perform RE tasks and interact with stakeholders. However, existing research primarily focuses on showcasing the capabilities of these tools rather than their design and evaluation in RE-specific contexts. This limits our understanding of their practical value and hinders broader adoption. To address this gap, we propose a reference model to guide the design, use, and evaluation of intelligent RE agents. Our work introduces new RE use cases, along with evaluation metrics for intelligent RE agents. We present a study design to support systematic development and share early findings demonstrating the feasibility of our approach. The use cases show how agents can add value for RE practitioners, while our synthesized catalogue supports tool evaluation. Finally, our analysis of commercial agents reveals that these tools already support certain aspects of the envisioned RE use cases.

Software products have been at the heart of the explosion of social computing technologies, but they are also implicated in facilitating online violence against women and girls. In this context, there is an acute need for responsible approaches to software engineering that help minimise the risks of online harms through effective consideration of how harms could arise in the context of a particular software product and selection of appropriate mitigations to minimise the likelihood and impact of these harms. In this paper, we highlight some of the key challenges faced by software engineering teams who want to ensure that software products serve the needs of users, while also protecting these users from harms. To address these challenges, we propose that there is an essential prerequisite to develop and maintain a knowledge model of online harms. Such a knowledge model could be used by software engineers to explore how online harms link to relevant policy and regulation, the actors involved in online harm incidents; other relevant contextual factors; and product features. We propose a meta-model for capturing this knowledge and identify some key research challenges to address for software development teams to practice responsible software engineering for online safety.

This study explores the application of predictive analytics in evaluating player performance in the National Basketball Association (NBA), focusing on rebounds per game (REB), an essential component for better performance and results in basketball. The research employs a comparative analysis of machine learning (ML) models by leveraging a detailed NBA dataset. A key novelty lies in integrating advanced hyperparameter tuning and feature selection, enabling these models to capture complex relationships within the dataset. The Gradient Boosting Regressor demonstrated superior predictive performance, achieving an R² score of 0.8749 after tuning, with Linear Regression following closely at 0.8668. This study also highlights the importance of model interpretability and scalability, emphasizing the balance between predictive accuracy and usability for real-world decision-making. By offering actionable insights for optimizing player strategies and team performance, this research contributes to the growing body of knowledge in data-driven sports analytics and paves the way for more advanced applications in professional basketball management.

Deepfakes are one of the most recent developments in misinformation technology and are capable of superimposing one person's face onto another in video format. The potential of this technology to defame and cause harm is clear. However, despite the grave concerns expressed about deepfakes, these concerns are rarely accompanied with empirical evidence. We present a scoping review of the existing empirical studies that aim to investigate the effects of viewing deepfakes on people's beliefs, memories, and behaviour. Five databases were searched, producing an initial sample of 2004 papers, from which 22 relevant papers were identified, varying in methodology and research methods used. Overall, we found that the early studies on this topic have often produced inconclusive findings regarding the existence of uniquely persuasive or convincing effects of deepfake exposure. Moreover, many experiments demonstrated poor methodology and did not include a non-deepfake comparator (e.g., text-based misinformation). We conclude that speculation and scare mongering about dystopian uses of deepfake technologies has far outpaced experimental research that assess these harms. We close by offering insights on how to conduct improved empirical work in this area.

Software engineering skills are broad and varied, encompassing not only technical abilities, but cognitive and social dimensions as well. Previous research establishes soft skills as being central for software engineering, e.g., teamwork, communication, and problem solving, but the relationship between these skills and how higher education prepares future software engineers for the workplace is unclear. These programmes should teach students the soft skills required for their careers, but a better understanding of which skills are valued and taught is needed. The perceptions of undergraduate alumni about the value of soft skills and where they were developed are reported on in this preregistered study of computer science and psychology alumni (using the latter as a contrast group). Replicating previous conclusions, computer science alumni rated problem solving, communication, and teamwork as amongst the most important skills, but when examining where these skills originate from, problem solving was strongly associated with education, and teamwork was strongly linked with employment. Communication is weakly associated with education, with influence from employment sources too. Alumni working in security-related careers highlighted that organisational skills were important more so than for the alumni population more broadly. We make recommendations for prioritisation of skill development based on the skills reported by the alumni. We suggest potential ways to encourage learning opportunities that promote scenarios students will find familiar during their careers, better preparing students for the needs of the workplace.

Robust initialisation has shown to greatly improve the performance of genetic programming on a wide variety of tasks. Many of these have been adapted to work with grammatical evolution, with varying success. We are the first to examine the effectiveness of some of the most popular grammatical evolution initialisation techniques using structured grammatical evolution. Namely, we investigate sensible initialisation and probabilistic tree creation 2, as well as the standard initialisation procedure used in structured grammatical evolution, grow.

We draw on the concept of episodic volunteering (EV) from the general volunteering literature to identify practices for managing EV in free/libre/open source software (FLOSS) communities. Infrequent but ongoing participation is widespread, but the practices that community managers are using to manage EV, and their concerns about EV, have not been previously documented. We conducted a policy Delphi study involving 24 FLOSS community managers from 22 different communities. Our panel identified 16 concerns related to managing EV in FLOSS, which we ranked by prevalence. We also describe 65 practices for managing EV in FLOSS. Almost three-quarters of these practices are used by at least three community managers. We report these practices using a systematic presentation that includes context, relationships between practices, and concerns that they address. These findings provide a coherent framework that can help FLOSS community managers to better manage episodic contributors.

Automation systems involve a range of cyber-physical system components such as sensors, actuators, control equipment, machines, robots, AGVs, etc. Seamless interoperability among these entities is a significant challenge. A well-designed Industrial Internet of Things (IIoT) platform at the network edge can offer several services by acting as a transformation engine between these field devices and various enterprise applications. We consider the EdgeX Foundry platform as such an IIoT middleware, discuss how EdgeX can provide ready-to-use integration of IoT devices, and show how we connect it with a low-code XMDD coordination layer that interfaces with EdgeX microservices through a Native DSL mechanism. We consider this technology landscape from the point of view of a building automation system example that supports high reconfigurability and security. We show how to produce all the essential elements of a complex Web based application to control the considered building systems. We demonstrate various features of the application's data and process models, how DSLs play a role at various levels, and how to add security capabilities that go beyond the cross-layer concerns and mechanisms offered by EdgeX. To this end, we introduce a declarative policy layer to be implemented using the open source ADD-Lib in form of an additional DSL for Attribute Based Encryption, with the aim of further enriching the capabilities around the EdgeX platform.