• Korean Management Science Review
• /
• v.32 no.2
• /
• pp.111-124
• /
• 2015

In the management literature, it has been widely accepted among both researchers and practitioners that the level of team creativity is significantly related to the organizational performance. Besides, researchers posited with confidence that team diversity and task difficulty would affect team creativity through team members' activities of exploration and exploitation. However, empirical approaches to proving this belief suffered from lack of real data and proper methods as well. To tackle the research void like this, we propose a multi-agent simulation (MAS) mechanism. By adopting a set of parameters which validity were proven in the related literature, we conducted a series of experiments in the environment of the MAS platform named NetLogo. There sults suggest that managers can differentiate team diversity strategies according to task difficulty. In the case of a difficult task, managers need to increase team diversity so that their teams can maximize team creativity through rigorous exploration and exploitation. It is desirable to maintain an average level of team diversity when performing an easy task.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.5
• /
• pp.1238-1259
• /
• 2024

Previous studies on joint optimization of computation offloading and service caching policies in Mobile Edge Computing (MEC) have often neglected the impact of dependency-aware subtasks, edge server resource constraints, and multiple users on policy formulation. To remedy this deficiency, this paper proposes a many-objective joint optimization dependency-aware task offloading and service caching model (MaJDTOSC). MaJDTOSC considers the impact of dependencies between subtasks on the joint optimization problem of task offloading and service caching in multi-user, resource-constrained MEC scenarios, and takes the task completion time, energy consumption, subtask hit rate, load variability, and storage resource utilization as optimization objectives. Meanwhile, in order to better solve MaJDTOSC, a many-objective evolutionary algorithm TSMSNSGAIII based on a three-stage mating selection strategy is proposed. Simulation results show that TSMSNSGAIII exhibits an excellent and stable performance in solving MaJDTOSC with different number of users setting and can converge faster. Therefore, it is believed that TSMSNSGAIII can provide appropriate sub-task offloading and service caching strategies in multi-user and resource-constrained MEC scenarios, which can greatly improve the system offloading efficiency and enhance the user experience.

• ETRI Journal
• /
• v.45 no.1
• /
• pp.93-104
• /
• 2023

Coreference resolution is a task in discourse analysis that links several headwords used in any document object. We suggest pointer networks-based coreference resolution for Korean using multi-task learning (MTL) with an attention mechanism for a hierarchical structure. As Korean is a head-final language, the head can easily be found. Our model learns the distribution by referring to the same entity position and utilizes a pointer network to conduct coreference resolution depending on the input headword. As the input is a document, the input sequence is very long. Thus, the core idea is to learn the word- and sentence-level distributions in parallel with MTL, while using a shared representation to address the long sequence problem. The suggested technique is used to generate word representations for Korean based on contextual information using pre-trained language models for Korean. In the same experimental conditions, our model performed roughly 1.8% better on CoNLL F1 than previous research without hierarchical structure.

• Journal of Institute of Control, Robotics and Systems
• /
• v.5 no.2
• /
• pp.145-150
• /
• 1999

A real-time control system, composed of the controlled processor and the controller computer(s), may have a variety of task types, some of which have tight timing-constraints in generating the correct control input. The maximum period of those task failures tolerable by the system is called the hard deadline, which depends on not only fault characteristics but also task characteristics. In the paper, we extend a method deriving the hard deadline in LTI system executing single task. An algorithm to combine the deadlines of all the elementary tasks in the same operation-mode is proposed to derive the hard deadline of the entire system. For the end, we modify the state equation for the task to capture the effects of task failures (delays in producing correct values) and inter-correlation. We also classify the type of executing the tasks according to operation modes associated with relative importance of correlated levels among tasks, into series, parallel, and cascade modes. Some examples are presented to demonstrate the effectiveness of the proposed methods.

• Journal of Information Processing Systems
• /
• v.19 no.4
• /
• pp.450-464
• /
• 2023

To address the problems of large system overhead and low timeliness when dealing with task scheduling in mobile edge cloud computing, a task scheduling and resource management strategy for edge cloud computing based on an improved genetic algorithm was proposed. First, a user task scheduling system model based on edge cloud computing was constructed using the Shannon theorem, including calculation, communication, and network models. In addition, a multi-objective optimization model, including delay and energy consumption, was constructed to minimize the sum of two weights. Finally, the selection, crossover, and mutation operations of the genetic algorithm were improved using the best reservation selection algorithm and normal distribution crossover operator. Furthermore, an improved legacy algorithm was selected to deal with the multi-objective problem and acquire the optimal solution, that is, the best computing task scheduling scheme. The experimental analysis of the proposed strategy based on the MATLAB simulation platform shows that its energy loss does not exceed 50 J, and the time delay is 23.2 ms, which are better than those of other comparison strategies.

• Proceedings of the IEEK Conference
• /
• 1998.10a
• /
• pp.511-514
• /
• 1998

This paper proposes efficient methods that integrate and execute local plan rules of task agents in a multi-agent environment. In these methods, each agent's plan rules are represented in a network structure, and these networks are then collected by a single task agent to build a integrated domain network, which is exploited to achieve the goal. Agent problem solving by using the domain network enables a concurrent execution of plan rules that are sequential in nature.

• Journal of Institute of Control, Robotics and Systems
• /
• v.7 no.11
• /
• pp.927-932
• /
• 2001

When people write a character with a pen stably, proper compliance planning is necessary. In this paper, after investigating the property of character writing task, we propose a fundamental grasp model for character writing and also analyze compliance characteristics for effective character writing using multi-fingered hands. For this, the general stiffness relation of multi-fingered hand is firstly described. Next, we investigate the grasp configurations for grasping a pen and then, we analyze the conditions of the specified stiffness matrix in the operational space to successfully and more effectively achieve the given character writing task. Through the analysis, an effective grasp modeling for successful character writing is shown. And also, we conclude that the operational compliance characteristics should be properly planned for character writing, stably and precisely.

• Journal of Information Processing Systems
• /
• v.17 no.5
• /
• pp.867-878
• /
• 2021

In context awareness and user intention tasks, dataset construction is expensive because specific domain data are required. Although pretraining with a large corpus can effectively resolve the issue of lack of data, it ignores domain knowledge. Herein, we concentrate on data domain knowledge while addressing data scarcity and accordingly propose a multi-channel long short-term memory (LSTM). Because multi-channel LSTM integrates pretrained vectors such as task and general knowledge, it effectively prevents catastrophic forgetting between vectors of task and general knowledge to represent the context as a set of features. To evaluate the proposed model with reference to the baseline model, which is a single-channel LSTM, we performed two tasks: voice phishing with context awareness and movie review sentiment classification. The results verified that multi-channel LSTM outperforms single-channel LSTM in both tasks. We further experimented on different multi-channel LSTMs depending on the domain and data size of general knowledge in the model and confirmed that the effect of multi-channel LSTM integrating the two types of knowledge from downstream task data and raw data to overcome the lack of data.

• Journal of Korea Society of Industrial Information Systems
• /
• v.27 no.2
• /
• pp.61-69
• /
• 2022

A multi-agent system is a system that aims at achieving the best-coordinated decision based on each agent's local decision. In this paper, we consider a multi agent-multi task assignment problem. Each agent is assigned to only one task and there is a completion probability for performing. The objective is to determine an assignment that maximizes the sum of the completion probabilities for all tasks. The problem, expressed as a non-linear objective function and combinatorial optimization, is NP-hard. It is necessary to design an effective and efficient solution methodology. This paper presents an approximation algorithm using submodularity, which means a marginal gain diminishing, and demonstrates the scalability and robustness of the algorithm in theoretical and experimental ways.

• ETRI Journal
• /
• v.46 no.1
• /
• pp.82-95
• /
• 2024

As automated essay scoring (AES) has progressed from handcrafted techniques to deep learning, holistic scoring capabilities have merged. However, specific trait assessment remains a challenge because of the limited depth of earlier methods in modeling dual assessments for holistic and multi-trait tasks. To overcome this challenge, we explore providing comprehensive feedback while modeling the interconnections between holistic and trait representations. We introduce the DualBERT-Trans-CNN model, which combines transformer-based representations with a novel dual-scale bidirectional encoder representations from transformers (BERT) encoding approach at the document-level. By explicitly leveraging multi-trait representations in a multi-task learning (MTL) framework, our DualBERT-Trans-CNN emphasizes the interrelation between holistic and trait-based score predictions, aiming for improved accuracy. For validation, we conducted extensive tests on the ASAP++ and TOEFL11 datasets. Against models of the same MTL setting, ours showed a 2.0% increase in its holistic score. Additionally, compared with single-task learning (STL) models, ours demonstrated a 3.6% enhancement in average multi-trait performance on the ASAP++ dataset.