Multimodal emotion recognition is essential in affective computing, as it enables a more accurate and comprehensive understanding of human emotions by integrating diverse data modalities. However, current approaches still face key challenges, including the difficulty of handling unaligned multimodal inputs, limited ability to model long-term dependencies, and insufficient attention to relationships among emotional labels. To address these issues, this paper introduces a unified framework that combines a Pseudo-Alignment Algorithm (PAA) for processing unaligned data, a Multimodal Data Interaction Process (MDIP) for fusing text, audio, and video while preserving long-term contextual information, and a Deep Reinforcement Learning-based Emotion Detection (DRLED) model for exploring inter-emotional dependencies. Experiments conducted on the IEMOCAP benchmark dataset demonstrate that the proposed approach achieves strong emotion recognition performance without relying on pre-aligned multimodal data, highlighting its effectiveness and robustness in real-world scenarios.
Lateral walking exercise is beneficial for the hip abductor enhancement. Accurate gait recognition and continuous hip joint angle prediction are essential for the control of exoskeletons. We propose a dual-task learning framework, the “Twin Brother” model, which fuses convolutional neural network (CNN), long short-term memory (LSTM), neural networks (NNs), and the squeezing-elicited attention mechanism to classify the lateral gait stage and estimate the hip angle from electromyography (EMG) signals. The EMG signals of 6 muscles from 10 subjects during lateral walking were collected. Four gait phases were recognized, and the hip angles of both legs were continuously estimated. The sliding window length of 250 ms and the sliding increment of 3 ms were determined by the requirements of response time and recognition accuracy of the real-time system. We compared the performance of CNN-LSTM, CNN, LSTM, support vector machine, NN, K-nearest neighbor, and the “Twin Brother” models. The “Twin Brother” model achieved a recognition accuracy (mean ± SD) of 98.81% ± 0.14%. The model’s predicted root mean square error (RMSE) for the left and right hip angles are 0.9183° ± 0.024° and 1.0511° ± 0.027°, respectively, where the R2 are 0.9853 ± 0.006 and 0.9808 ± 0.008. The accuracy of recognition and estimation are both better than comparative models. For gait phase percentage prediction, RMSE and R2 predicted by the model can reach 0.152° ± 0.014° and 0.986 ± 0.011, respectively. These results demonstrate that the method can be applied to lateral walking gait recognition and hip joint angle prediction.
Il presente lavoro si propone di elaborare alcuni spunti per una riflessione sull’evoluzione degli strumenti di risoluzione delle controversie online (ODR) alla luce delle preoccupanti segnalazioni in relazione alla c.d. frammentazione di Internet, ossia quel fenomeno, denunciato dai fautori della neutralità della rete, che si sta verificando in rete a vari livelli, sia con riguardo all’infrastruttura che ai contenuti e che si ritiene possa mettere in serio pericolo la possibilità di un reale accesso universale e indiscriminato ai servizi offerti in rete. Considerando che la costellazione degli strumenti e piattaforme ODR appartiene al mondo privato, emerge fortemente il rischio che si possano imporre condizionamenti, anche in modo surrettizio, rispetto alla promozione dell’uso e in definitiva all’efficacia deflattiva del ricorso alla giurisdizione pubblica, soprattutto a danno dei soggetti più deboli.
What do we miss about the daily operations of platform power, and about power dynamics in the gig economy more broadly, when focusing on algorithmic management as the primary source of subordination and precarity in the workplace? Drawing on a five-year research project investigating platform-based food delivery and domestic cleaning in Amsterdam, Berlin, and New York City, this paper advances the argument that in order to understand the situated and contingent nature of platform power in the gig economy we should examine how gig workers manage risk. While a handful of studies have explicitly addressed this topic, we still know little about how socioeconomic stratification within gig workforces mediates workers’ vulnerability to various kinds of risk, as well as their susceptibility to platform power. In response, the paper develops a “platform-adjacent” approach that situates gig work within people’s larger work and life trajectories. It demonstrates how gig platforms can become both a resource for risk management and a new source of risk, depending on the complex interaction between a platform’s labour management strategies on the one hand and the mix of support structures and dependencies in a worker’s life on the other. Ultimately, the paper offers a more nuanced and comprehensive understanding of how gig platforms become integrated into people’s everyday life and how platform power is articulated and negotiated over time.
Alberto Benayas, Miguel Angel Sicilia, Marçal Mora-Cantallops
Conversational Agents (CA) are increasingly being deployed by organizations to provide round-the-clock support and to increase customer satisfaction. All CA have one thing in common despite the differences in their design: they need to be trained with users’ intents and corresponding training sentences. Access to proper data with acceptable coverage of intents and training sentences is a big challenge in CA deployment. Even with the access to the past conversations, the process of discovering intents and training sentences manually is not time and cost-effective. Here, an end to end automated framework that can discover intents and their training sentences in conversation logs to generate labeled data sets for training intent models is presented. The framework proposes different feature engineering techniques and leverages dimensionality reduction methods to assemble the features, then applies a density-based clustering algorithm iteratively to mine even the least common intents. Finally, the clustering results are automatically labeled by the final algorithm.
Relevance. The Monopoly game is a popular board game with a simple concept that is relevant as a tool for enterprises and entrepreneurs in the system of teaching the basics of business, as a tool for studying strategic economics and justifying business strategies. Goals. Substantiation of the importance of the Monopoly game model as a business strategy tool for developing skills in strategic thinking, negotiation, risk management and efficient resource allocation. Exploring the similarities between concepts in the game and real economic concepts, including monopoly, market competition, pricing strategies, and business cycles. Evaluation of the effectiveness of the use of “Monopoly” as an educational tool for teaching the concepts of strategic economics. Methodology. Analysis, synthesis, observation, comparative and statistical methods. Conclusions. The business game “Monopoly” is a useful tool in the system of teaching economic concepts, raising the level of knowledge of the real economy. The use of the game model “Monopoly” in the educational process develops critical thinking and the skills of effective management decisions focused on business development.
This paper discusses new challenges in learning informatics from an interdisciplinary perspective. As topics of learning informatics must cover very wide ranges, this paper will provide an overview with a combination of several ideas and applications. From a methodological point of view, the challenges include the concepts of 'knowledge networks', 'learner personas', and 'visualization of the experience'. From the viewpoints of educational applications, they contain the concepts of 'class design', 'peer review' and 'case learning'. The introduction of various concepts from fields such as computer sciences, marketing research, and design thinking of soft systems domains, whose areas are far from education applications, would add new insights to learning informatics research.
Michael Stephan, Lorenzo Servadei, José Arjona-Medina
et al.
Multi-target tracking with radars is a highly challenging problem due to detection artifacts, sensor noise, and interference sources. The traditional signal processing chain is, therefore, a complex combination of various algorithms with several tunable tracking-parameters. Usually, these are initially set by engineers and are independent of the scene tracked. For this reason, they are often non-optimal and generate poorly performing tracking. In this context, scene-adaptive radar processing refers to algorithms that can sense, understand and learn information related to detected targets as well as the environment and adapt its tracking-parameters to optimize the desired goal. In this paper, we propose a Deep Reinforcement Learning framework that guides the scene-adaptive choice of radar tracking-parameters towards an improved performance on multi-target tracking.
Eugene Aleksandrov, Tetiana Aleksandrova, Iryna Kostianyk
et al.
The problem of choosing the values of the variable parameters of the digital stabilizer of the cosmic stage of a carrier rocket with a liquid-propellant jet engine and an onboard digital computer in the stabilization loop, which ensures stable movement of the stage along the entire active section of the flight trajectory, is considered. The effect of the stabilizer quantization period on the stability region of a closed-loop stabilization system is considered. It is recommended to choose the intersection of stability regions corresponding to uniformly distributed moments of time along the active section of the stage flight trajectory as acceptable values for the variable parameters of the stabilizer of a non-stationary stabilization object.
Grid computing is a computational paradigm that emerged to handle the increasing demand for computational resources. Several metaheuristics methods have been applied to tackle the grid task scheduling problem. These metaheuristics generally generate good but not optimal task schedules. The aim of this paper is to design and implement a grid task scheduling mechanism to map clients’ tasks to available resources in order to finish the submitted tasks within the optimal execution time. The paper proposes an enhanced time shared metaheuristics mechanism based on Firefly Algorithm to improve the grid job scheduling process. The proposed mechanism utilizes the Smallest Position Value (SPV) technique to handle the scheduling problem as permutations. Experiments using simulations and real workload traces were conducted to study the performance of the proposed enhanced time shared metaheuristic scheduling mechanism. Empirical results revealed that the proposed timed shared metaheuristic algorithm can efficiently reduce the makespan time to 1851 compared with 3482, 3185 for Tabu search and genetic algorithm, respectively.
In this paper, we put forth an evolve-then-correct reduced order modeling approach that combines intrusive and nonintrusive models to take hidden physical processes into account. Specifically, we split the underlying dynamics into known and unknown components. In the known part, we first utilize an intrusive Galerkin method projected on a set of basis functions obtained by proper orthogonal decomposition. We then present two variants of correction formula based on the assumption that the observed data are a manifestation of all relevant processes. The first method uses a standard least-squares regression with a quadratic approximation and requires solving a rank-deficient linear system, while the second approach employs a recurrent neural network emulator to account for the correction term. We further enhance our approach by using an orthonormality conforming basis interpolation approach on a Grassmannian manifold to address off-design conditions. The proposed framework is illustrated here with the application of two-dimensional co-rotating vortex simulations under modeling uncertainty. The results demonstrate highly accurate predictions underlining the effectiveness of the evolve-then-correct approach toward real-time simulations, where the full process model is not known a priori.
هدف: هدف این پژوهش ارائه تحلیلی از تحقیقات تحلیل محتوایی انجام شده در حوزه علم اطلاعات و دانششناسی ایران بود.
روش: پژوهش حاضر از نوع تحقیقات کاربردی است که با روش تحلیل محتوا انجام شد.
یافته ها: نشان داد که در بازه زمانی مورد مطالعه پژوهشگران به طور کلی ۴۴ پژوهش تحلیل محتوا منتشر شده است که از آنها، ۹ عنوان در قالب پایاننامه، ۳۱ عنوان در قالب مقاله، ۴ عنوان در قالب کتاب و یا بخشی از یک کتاب منتشر شده است. ۱۶ عنوان، مقالهها (۱۲ عنوان مقالههای مجلات و ۴ عنوان مقاله های همایش ها)، ۱۰ عنوان پایاننامه ها، ۱۲ عنوان کتاب ها و ۶ عنوان کلیه متون حوزه را تحلیل کرده اند. ۲۱ عنوان (۷/۴۷ درصد) بهصورت چند نویسنده ای و ۲۳ عنوان (۳/۵۲ عنوان) بهصورت تک نویسنده ای منتشر شده است. از نظر زمانی ۲ عنوان در دهه ۱۳۶۰، ۱۱ عنوان در دهه ۱۳۷۰، ۱۵ عنوان در دهه ۱۳۸۰ و ۱۶ عنوان در دهه ۱۳۹۰ منتشر شده اند.
نتیجه گیری: تحقیقات تحلیل محتوا ضمن مشخص کردن خلأهای پژوهشی، محققان را بهسوی آنها خلاء ها راهنمایی میکنند. علی رغم کاربرد بیشتر این روش در حوزههای مختلف، در حوزه علم اطلاعات و دانششناسی چندان مورد توجه نبوده و لازم است بیشتر مورد عنایت قرار گیرد.
This article describes a formal proof of the Kepler conjecture on dense sphere packings in a combination of the HOL Light and Isabelle proof assistants. This paper constitutes the official published account of the now completed Flyspeck project.
Case Based Reasoning (CBR) is a computer system that used for reasoning old knowledge to solve new problems. It works by looking at the closest old case to the new case. This research attempts to establish a system of CBR for diagnosing heart disease. The diagnosis process is done by inserting new cases containing symptoms into the system, then the similarity value calculation between cases uses the nearest neighbor method similarity, minkowski distance similarity and euclidean distance similarity.
Case taken is the case with the highest similarity value. If a case does not succeed in the diagnosis or threshold <0.80, the case will be revised by experts. Revised successful cases are stored to add the systemknowledge. Method with the best diagnostic result accuracy will be used in building the CBR system for heart disease diagnosis.
The test results using medical records data validated by expert indicate that the system is able to recognize diseases heart using nearest neighbor similarity method, minskowski distance similarity and euclidean distance similarity correctly respectively of 100%. Using nearest neighbor get accuracy of 86.21%, minkowski 100%, and euclidean 94.83%
This paper, Change Requires Change: will relate that bridging the gap between education: of what we teach and training: of what industry looks for in prepared skills for students, needs to be relevant to today's situations.<br><br> We need to re-evaluate traditional industry academic partnerships which have been relatively successful including; internships, work-study programs, curriculum advisory boards, guest lectures and capstone courses, to identify gaps and opportunities for what is needed to support our future.<br><br> Do we want to continue with the status-quo or enhance education? Should we be cognizant of emerging trends? What could be the implications on changing academic-industry partnerships? How can we improve? This paper proposes several new approaches to academics and industry practitioner's towards greater successful collaborations towards student preparation.