Search Results (44)

The widespread adoption of cloud computing has amplified the demand for electric power. It is strategically important to address the limitations of reliable sources and sustainability of power. Research and investment in data centres and power infrastructure are therefore critically important for our digital economy. A novel heuristic for the minimisation of energy consumption in cloud computing is presented. It draws similarities to the concept of “network slices”, in which an orchestrator enables multiplexing to reduce the network “churn” often associated with significant losses of energy consumption. The novel network slicing–ring fencing ratio is a heuristic calculated through an iterative procedure for the reduction in cloud energy consumption. Simulation results show how the non-convex equation optimises power by reducing energy from 10,680 kJ to 912 kJ, which is a 91.46% efficiency gain. In comparison, the Heuristic AUGMENT Non-Convex algorithm (HA-NC, by Hossain and Ansari) reported a 312.74% increase in energy consumption from 2464 kJ to 10,168 kJ, while the Priority Selection Offloading algorithm (PSO, by Anajemba et al.) also reported a 150% increase in energy consumption, from 10,738 kJ to 26,845 kJ. The proposed network slicing–ring fencing ratio is seen to successfully balance energy consumption and computing performance. We therefore think the novel approach could be of interest to network architects and cloud operators. Full article

User churn in online games refers to players becoming inactive for an extended period. Even a small increase in churn can lead to significant revenue loss, making churn prediction crucial for sustaining long-term player engagement. Although user churn prediction has been extensively studied, most existing approaches either ignore non-login periods or treat all inactivity uniformly, overlooking key behavioral differences. This study addresses this gap by categorizing non-login periods into three types, as follows: inactivity due to new or dormant users, genuine loss of interest, and temporary inaccessibility caused by external factors. These periods are treated as either non-existent or missing data and imputed using techniques such as mean or mode substitution, linear interpolation, and multiple imputation by chained equations (MICE). MICE was selected due to its ability to impute missing values more robustly by considering multivariate relationships. A random forest (RF) classifier, chosen for its interpretability and robustness to incomplete data, serves as the primary prediction model. Additionally, classifier chains are used to capture label dependencies, and principal component analysis (PCA) is applied to reduce dimensionality and mitigate overfitting. Experiments on real-world MMORPG data show that our approach improves predictive accuracy, achieving a micro-averaged AUC of above 0.92 and a weighted F1 score exceeding 0.70. These findings suggest that our approach improves churn prediction and offers actionable insights for supporting personalized player retention strategies. Full article

In this paper, we propose a methodology designed to deliver actionable insights that help businesses retain customers. While Machine Learning (ML) techniques predict whether a customer is likely to churn, this alone is not enough. Explainable Artificial Intelligence (XAI) methods, such as SHapley Additive Explanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME), highlight the features influencing the prediction, but businesses need strategies to prevent churn. Counterfactual (CF) explanations bridge this gap by identifying the minimal changes in the business–customer relationship that could shift an outcome from churn to retention, offering steps to enhance customer loyalty and reduce losses to competitors. These explanations might not fully align with business constraints; however, alternative scenarios can be developed to achieve the same objective. Among the six classifiers used to detect churn cases, the Balanced Random Forest classifier was selected for its superior performance, achieving the highest recall score of 0.72. After classification, Diverse Counterfactual Explanations with ML (DiCEML) through Mixed-Integer Linear Programming (MILP) is applied to obtain the required changes in the features, as well as in the range permitted by the business itself. We further apply DiCEML to uncover potential biases within the model, calculating the disparate impact of some features. Full article

This paper investigates the special phenomenon that the practical immersed depth of a spiral bevel gear as the driving gear under splash lubrication is significantly less than the static depth. To quantify the practical immersion depth, a computational fluid dynamics (CFD) approach integrated with image processing techniques is utilized to determine the dynamic immersion depth and the associated churning power loss. First, a theoretical method is developed to estimate the churning losses of the bevel gear by replacing the static immersion depth with the practical dynamic immersion depth. Subsequently, the CFD method, which incorporates the overset mesh technique and the volume-of-fluid (VOF) method, is employed to simulate the gear churning phenomenon. Meanwhile, the dynamic immersion depth is determined through image processing techniques that analyze the oil distribution characteristics in the splash-lubricated bevel gear. Finally, experimental results obtained from a dedicated lubrication test rig are favorably compared with the numerical results, confirming that the practical dynamic immersion depth is an accurate and effective parameter for calculating power losses. Full article

In the current energy landscape, efficiency is a critical topic. Therefore, even in the case of geared transmissions, it is essential to predict and calculate power losses as accurately as possible from the design phase. There are mainly three categories of losses in a gear unit: friction—the power losses due to the contact between teeth in rotation on the one hand and the seals with the spindles on the other hand; churning—the power losses generated by the air–lubricant mixture compression around teeth roots during rotation; and windage—the power losses due to the teeth aerodynamic trail in the air–lubricant mixture. While the first two categories of losses are intensively studied in the literature, the papers focusing on windage power losses are less representative. An estimation of windage power losses can be performed by numerical simulation, and the accuracy of the results depends on the mesh density and the available computing power. The present study discusses the influence of meshing on the windage torque of an orthogonal face gear immersed in air and compares numerical results generated by SolidWorks 2025 Flow Simulation software with experimental data measured on a test rig. Full article

Customers are the cornerstone of business success across industries. Companies invest significant resources in acquiring new customers and, more importantly, retaining existing ones. However, customer churn remains a major challenge, leading to substantial financial losses. Addressing this issue requires a deep understanding of customers’ cognitive status and behaviours, as well as early signs of churn. Predictive and Machine Learning (ML)-based analysis, when trained with appropriate features indicative of customer behaviour and cognitive status, can be highly effective in mitigating churn. A robust ML-driven churn analysis depends on a well-developed feature engineering process. Traditional churn analysis studies have primarily relied on demographic, product usage, and revenue-based features, overlooking the valuable insights embedded in customer–company interactions. Recognizing the importance of domain knowledge and human expertise in feature engineering and building on our previous work, we propose the Customer Churn-related Knowledge Base (ChurnKB) to enhance feature engineering for churn prediction. ChurnKB utilizes textual data mining techniques such as Term Frequency-Inverse Document Frequency (TF-IDF), cosine similarity, regular expressions, word tokenization, and stemming to identify churn-related features within customer-generated content, including emails. To further enrich the structure of ChurnKB, we integrate Generative AI, specifically large language models, which offer flexibility in handling unstructured text and uncovering latent features, to identify and refine features related to customer cognitive status, emotions, and behaviours. Additionally, feedback loops are incorporated to validate and enhance the effectiveness of ChurnKB.Integrating knowledge-based features into machine learning models (e.g., Random Forest, Logistic Regression, Multilayer Perceptron, and XGBoost) improves predictive performance of ML models compared to the baseline, with XGBoost’s F1 score increasing from 0.5752 to 0.7891. Beyond churn prediction, this approach potentially supports applications like personalized marketing, cyberbullying detection, hate speech identification, and mental health monitoring, demonstrating its broader impact on business intelligence and online safety. Full article

The occurrence of human resource churning results in financial, time and effort losses for organisations, which creates a problem for organisations that lose the most experienced human resources that they have invested in training. The human resources that leave organisations are known as churners. Churning is the costly, time-consuming and difficult process of replacing workers who have left voluntarily. Given the multiplicity of definitions attributed to the subject of churning, we follow the approach that human resource churning is a component of turnover, which is related to analysing the costs associated with voluntary departures. As a result of this problem, this entry was created with the aim of theoretically explaining the effects that the churning of human resources has on organisations of origin. In order to meet this objective, various topics are covered with the aim of characterising churners, their backgrounds and their aspirations, referring to the effect of the mobility of human resources on organisations, in other words, the effect of churning on organisations and the urgent need for action. Full article

In the dual-channel retail industry, excessive enthusiasm in offline retailers’ services often extends beyond the customer’s “interpersonal distance zone”, leading to psychological discomfort for customers and a subsequent loss of demand. This situation can trap retailers in a dilemma known as the “service trap”. To address this issue, we introduce the concept of the zone of service tolerance, which encompasses desired and adequate levels of service, into a dual-channel supply chain consisting of an online channel manufacturer and an offline retailer. We incorporate the zone of service tolerance into the demand function of the offline retailer and establish its profit function, a dynamic game theory to demonstrate the existence of a linkage mechanism between the optimal selling price and service level, providing the conditions for such a mechanism to exist. Additionally, we establish conditions for offline retailers to avoid over-servicing or under-servicing and consider the impacts of these conditions, and we reveal the stability conditions of the offline retailers’ service decisions. Our findings indicate that both over- and under-servicing can lead to customer churn. For newly launched products, offline retailers risk losing customers by adopting a sales strategy focused on high profits and moderate sales (under-servicing). Similarly, for products nearing removal from the shelves, they risk losing customers by adopting a sales strategy focused on low profits and high sales (over-servicing). Furthermore, under certain ranges for the service sensitivity factor, desired service, or adequate service, the optimal service provided by offline retailers remains robust regardless of the manufacturer’s optimal selling price. This greatly simplifies the offline retailer’s decision-making process regarding service levels, as they can directly focus on providing the desired service without factoring in the manufacturer’s pricing strategy. Full article

Subscription video-on-demand (SVOD) platforms face high churn rates and substantial revenue losses from SVOD content piracy, all of which limit their ability to invest in acquiring/creating content compelling enough to win and retain subscribers. Based on social exchange theory, this study argues that platforms can improve relationships with SVOD content users by offering tiered discounts in exchange for advertising/loyalty and by promoting anti-piracy messages with a prosocial (threatening) approach that emphasizes harm to filmmakers (punishment for pirates). We hypothesize that these incentives enhance subscription intention when the incentive specifications (advertising levels, loyalty levels, message approach, and message credibility) match the public’s heterogeneous dispositions (advertising attitude, loyalty attitude, justice sensitivity, and fear of punishment). In a survey on the intention to subscribe to a hypothetical new platform, we confirmed the hypothesized interactions for advertising-based discounts, loyalty-based discounts, and prosocial messages, but did not find support for threatening messages. Further exploration showed that the evaluation of platform content was much more influential than any other incentive and that tiered loyalty discounts had a remarkable capacity to enhance subscription intention. This study’s findings may help shape incentives that are more satisfying to users and ultimately more profitable for platforms. Full article

This study investigates the power losses of rolling element bearings (REBs) lubricated using an oil bath. Experimental tests conducted on two different deep-groove ball bearings (DGBBs) provide valuable insights into the behaviour of DGBBs under different oil levels, generating essential data for developing accurate models of power losses. Observations of the oil bath dynamics reveal the formation of an oil ring at high oil levels, as observed for planetary gear trains, leading to modifications in the oil flow behaviour. The experiments demonstrate that oil bath lubrication generates power losses comparable to injection lubrication when the oil level is low. However, as the oil level increases, so do the power losses due to increased drag within the bearing. This study presents a comprehensive model for calculating drag losses. The proposed drag power loss model accounts for variations in oil level and significantly improves loss predictions. A comparison of existing models with the experimental results shows good agreement for both bearings, demonstrating the effectiveness of the developed model in accounting for oil bath height in loss calculations. Full article

This paper presents comparable sets of the no-load power loss as a product of windage and churning behaviors of a family of various rotating parts (i.e., disc, spur gear, straight bevel gear, and orthogonal face gear). Experimental measurements were carried out under pure air only and under partial immersion in oil to qualify and quantify the windage and churning effects of no-load power losses of a family of spur, bevel, and face gears along with a representative disc as the baseline. Aiming at exploring the influence of gear teeth on the total no-load power losses, two different theoretical analytical approaches are introduced to account for the churning contributions, by which the total power losses are estimated. Both analytical approaches compare well with the experimental findings. Furthermore, a spatial intersecting cross-axis gear (e.g., straight bevel gear and orthogonal face gear) results in higher no-load power losses than that of a representative disc or a parallel-axes gear. The significance of gear teeth (gear vs. disc) on windage behavior is presented, as well as the gear windage effects on the churning phenomenon in a high-speed splash-lubricated gear. Full article

In order to clarify the effect of mixed lubrication methods on the oil flow and power loss of the gearbox, this study adopts a high-precision moving particle semi-implicit (MPS) method to investigate the lubrication of the gearbox under the joint influence of splash lubrication and oil injection lubrication. The accuracy of the numerical method to calculate the churning torque was verified by the constructed test rig. The effects of rotational speed, immersion depth, injection volume rate, and oil injection angle were analyzed and evaluated for lubrication. The results show that better lubrication can be achieved with relatively small churning torques by using a hybrid lubrication method. This provides some references for engineering applications of gearboxes. Full article

To investigate the lubrication characteristics in high-speed train gearboxes, a two-stage herringbone gearbox with an idle gear was analyzed. The lubricant flow and distribution were shown using the moving particle semi-implicit (MPS) method. A liquid film flow model was brought in to enhance the non-slip wall boundary conditions, enabling MPS to predict the film flow characteristics. This study investigates the influence of gear rotating speed, lubricant volume, and temperature on lubricant flow, liquid film distribution, lubrication state in the meshing zone, and churning power loss. The results indicate that lubrication characteristics depend on the splashing effect of rotating gears and lubricant fluidity. Increasing gear rotating speed and lubricant temperature can improve liquid film distribution on the inner wall, increase lubricant volume, and thus enhance film thickness. The lubricant particles in the meshing zone correlate positively with the gear rotating speed and lubricant volume, correlate negatively with a temperature above 20 °C, and decrease notably at low temperatures. Churning power loss mainly comes from the output gear. As lubricant volume and gear rotating speed increase, churning torque and power loss increase. Above 20 °C, viscosity decreases, reducing power loss; low temperatures lessen lubricant fluidity, reducing churning power loss. Full article

In order to accurately and efficiently analyze the distribution law and motion status of lubricating oil in the spiral bevel gearbox of the electric multiple unit (EMU), a high-fidelity 3D CFD model of the spiral bevel gearbox of the EMU was established for the first time. The moving particle semi-implicit method was used to visualize the lubricating-oil flow field distribution characteristics of the gearbox. The distribution characteristics of lubricating oil in the gearbox with varying gear rotation speeds, initial lubricating-oil volume levels and oil temperatures were analyzed. It was found that the initial lubricating-oil volume is the factor with the largest influence, while the influences of gear rotation speed and oil temperature are relatively small. By analyzing the churning loss under various simulation conditions, it was found that the churning loss is positively correlated with the gear rotation speed and initial oil volume, and is more affected by the initial oil volume. The churning loss is negatively correlated with the oil temperature, and both are nonlinear relationships. The proportion of churning loss related to the driven gear is higher compared to that of the driving gear. These results can provide a theoretical basis for the subsequent optimization of the gearbox. Full article

Reverse osmosis (RO) is known for the economic dewatering of dairy streams without any change in phase. At the household level, surplus milk is fermented and churned to obtain butter, which is subsequently heated to obtain clarified milk fat (ghee). The production of 1 kg ghee generates 15–20 kg sour buttermilk (SBM) as a by-product that is mostly drained. This causes a loss of milk solids and environmental pollution. The processing, preservation and valorization of SBM are quite challenging because of its low total solids (TS) and pH, poor heat stability and limited shelf life. This investigation aimed to transform SBM into a novel dried dairy ingredient. SBM was thermized, filtered, defatted and concentrated at 35 ± 1 °C, employing RO up to 3.62× (12.86%). The RO concentrate was subsequently converted into sour buttermilk powder (SBMP) by employing spray drying. SBMP was further characterized for its physicochemical, reconstitution and functional properties; rheological and morphological characteristics; and amino acid and fatty acid profiling, along with FTIR and XRD spectra. SBMP was “instant soluble-3 s” and exhibited excellent emulsion stability (80.70%), water binding capacity (4.34 g/g of protein), flowability (28.36°) and antioxidant properties. In nutshell, a process was developed for the valorization of sour buttermilk to a novel dairy ingredient by employing reverse osmosis and a spray-drying process. Full article