• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06d
• /
• pp.230-231
• /
• 2011

In this paper, we propose a rendezvous mechanism for cognitive radio networks. In this mechanism, no prior knowledge of wireless nodes is required and it is totally distributed. Node can simply choose one of two strategies to rendezvous with its neighbors. The main benefit of this mechanism is eliminating the use of common control channel and centralized controller.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.05a
• /
• pp.1535-1537
• /
• 2007

A design pattern is a general solution to a commonly occurring problem. Design patterns have proven highly effective in representing, transferring, and applying the design knowledge in many engineering disciplines. However, these patterns have not addressed sensor network specifically. With a growth of sensors and sensor networks, and considering their profound applicability, there is a crucial need to articulate ones experience of application development or deployment of sensor nodes in the form of design patterns to avoid the future mistakes. This paper discusses the same issue and show applicability of design patterns in sensor networks.

• The KIPS Transactions:PartD
• /
• v.18D no.6
• /
• pp.501-508
• /
• 2011

A knowledge map, which has been recently applied in various fields, is discovering characteristics hidden in a large amount of information and showing a tangible output to understand the meaning of the discovery. In this paper, we suggested a knowledge map for research trend analysis based on keyword-relation networks which are constructed by using a database of the domestic journal articles in the computer engineering field from 2000 through 2010. From that knowledge map, we could infer influential changes of a research topic related a specific keyword through examining the change of sizes of the connected components to which the keyword belongs in the keyword-relation networks. In addition, we observed that the size of the largest connected component in the keyword-relation networks is relatively small and groups of high-similarity keyword pairs are clustered in them by comparison with the random networks. This implies that the research field corresponding to the largest connected component is not so huge and many small-scale topics included in it are highly clustered and loosely-connected to each other. our proposed knowledge map can be considered as a approach for the research trend analysis while it is impossible to obtain those results by conventional approaches such as analyzing the frequency of an individual keyword.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.10
• /
• pp.3230-3255
• /
• 2022

Causality mining in NLP is a significant area of interest, which benefits in many daily life applications, including decision making, business risk management, question answering, future event prediction, scenario generation, and information retrieval. Mining those causalities was a challenging and open problem for the prior non-statistical and statistical techniques using web sources that required hand-crafted linguistics patterns for feature engineering, which were subject to domain knowledge and required much human effort. Those studies overlooked implicit, ambiguous, and heterogeneous causality and focused on explicit causality mining. In contrast to statistical and non-statistical approaches, we present Bidirectional Encoder Representations from Transformers (BERT) integrated with Multi-level Feature Networks (MFN) for causality recognition, called BERT+MFN for causality recognition in noisy and informal web datasets without human-designed features. In our model, MFN consists of a three-column knowledge-oriented network (TC-KN), bi-LSTM, and Relation Network (RN) that mine causality information at the segment level. BERT captures semantic features at the word level. We perform experiments on Alternative Lexicalization (AltLexes) datasets. The experimental outcomes show that our model outperforms baseline causality and text mining techniques.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.3
• /
• pp.160-165
• /
• 2022

Mutual distillation is a knowledge distillation method that guides a cohort of neural networks to learn cooperatively by transferring knowledge between them, without the help of a teacher network. This paper aims to confirm whether mutual distillation is also applicable to super-resolution networks. To this regard, we conduct experiments to apply mutual distillation to the discriminators of SRGANs and analyze the effect of mutual distillation on improving SRGAN's performance. As a result of the experiment, it was confirmed that SRGANs whose discriminators shared their knowledge through mutual distillation can produce super-resolution images enhanced in both quantitative and qualitative qualities.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.12
• /
• pp.3416-3435
• /
• 2023

Task offloading in vehicular networks is hot topic in the development of autonomous driving. In these scenarios, due to the role of vehicles and pedestrians, task characteristics are changing constantly. The classical deep learning algorithm always uses a pre-trained neural network to optimize task offloading, which leads to system performance degradation. Therefore, this paper proposes a neural network fine-tuning task offloading algorithm, combining with location prediction for pedestrians and vehicles by the Payne model of fluid dynamics and the car-following model, respectively. After the locations are predicted, characteristics of tasks can be obtained and the neural network will be fine-tuned. Finally, the proposed algorithm continuously predicts task characteristics and fine-tunes a neural network to maintain high system performance and meet low delay requirements. From the simulation results, compared with other algorithms, the proposed algorithm still guarantees a lower task offloading delay, especially when congestion occurs.

• Journal of Scientific & Technological Knowledge Infrastructure
• /
• s.14
• /
• pp.26-31
• /
• 2004

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.447-450
• /
• 2003

In this research, we propose an automatic knowledge acquisition and composite knowledge expression mechanism based on machine learning and relational database. Most of traditional approaches to develop a knowledge base and inference engine of expert systems were based on IF-THEN rules, AND-OR graph, Semantic networks, and Frame separately. However, there are some limitations such as automatic knowledge acquisition, complicate knowledge expression, expansibility of knowledge base, speed of inference, and hierarchies among rules. To overcome these limitations, many of researchers tried to develop an automatic knowledge acquisition, composite knowledge expression, and fast inference method. As a result, the adaptability of the expert systems was improved rapidly. Nonetheless, they didn't suggest a hybrid and generalized solution to support the entire process of development of expert systems. Our proposed mechanism has five advantages empirically. First, it could extract the specific domain knowledge from incomplete database based on machine learning algorithm. Second, this mechanism could reduce the number of rules efficiently according to the rule extraction mechanism used in machine learning. Third, our proposed mechanism could expand the knowledge base unlimitedly by using relational database. Fourth, the backward inference engine developed in this study, could manipulate the knowledge base stored in relational database rapidly. Therefore, the speed of inference is faster than traditional text -oriented inference mechanism. Fifth, our composite knowledge expression mechanism could reflect the traditional knowledge expression method such as IF-THEN rules, AND-OR graph, and Relationship matrix simultaneously. To validate the inference ability of our system, a real data set was adopted from a clinical diagnosis classifying the dermatology disease.

• Korean Journal of Cognitive Science
• /
• v.10 no.4
• /
• pp.71-83
• /
• 1999

Conventional symbolic inference systems lack flexibility because they do not well reflect flexible semantic structure of knowledge and use symbolic logic for their basic inference mechanism. For solving this problem. we have recently proposed the 'Connectionist Semantic Network(CSN)' as a model for flexible knowledge representation and inference based on neural networks. The CSN is capable of carrying out both approximate reasoning and commonsense reasoning based on similarity and association. However. we have difficulties in representing general and structured high-level knowledge and variable binding using the connectionist framework of the CSN. In this paper. we propose a hybrid system called SymCSN(Symbolic CSN) that combines a symbolic module for representing general and structured high-level knowledge and a connectionist module for representing and learning low-level semantic structure Simulation results show that the SymCSN is a plausible model for human-like flexible knowledge representation and inference.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.54 no.5
• /
• pp.35-41
• /
• 2017

In this paper, we analyze the performance of the recently introduced Hint-knowledge distillation (KD) training approach based on the teacher-student framework for knowledge distillation and knowledge transfer. As a deep neural network (DNN) considered in this paper, the deep residual network (ResNet), which is currently regarded as the latest DNN, is used for the teacher-student framework. Therefore, when implementing the Hint-KD training, we investigate the impact on the weight of KD information based on the soften factor in terms of classification accuracy using the widely used open deep learning frameworks, Caffe. As a results, it can be seen that the recognition accuracy of the student model is improved when the fixed value of the KD information is maintained rather than the gradual decrease of the KD information during training.