• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.5
• /
• pp.801-808
• /
• 2022

Research on how to embed knowledge in large-scale Linked Data and apply neural network models for entity matching is relatively scarce. The most fundamental problem with this is that different labels lead to lexical heterogeneity. In this paper, we propose an extended GCN (Graph Convolutional Network) model that combines re-align structure to solve this lexical heterogeneity problem. The proposed model improved the performance by 53% and 40%, respectively, compared to the existing embedded-based MTransE and BootEA models, and improved the performance by 5.1% compared to the GCN-based RDGCN model.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.5
• /
• pp.239-245
• /
• 2006

Time-reversal processing (TRP) has been shown as an effective way to focus in both time and space. The temporal focusing properties have been used extensively in underwater acoustics communications. Recently. adaptive time-reversal processing (ATRP) was applied to the simultaneous multiple focusing in an ocean waveguide. In this study. multiple focusing with ATRP is extended to the underwater acoustic communication algorithm for multiple receiving locations. The developed algorithm is applied to the underwater acoustic communication to show, via simulation and real data, that the simultaneous self-equalization at multiple receiving locations is achieved.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.9
• /
• pp.363-372
• /
• 2019

A sequence refers to data in which the order exists on the two items, and purchase transaction data in which the products purchased by one customer are listed is one of the representative sequence data. In general, all goods have a product taxonomy, such as category/ sub-category/ sub-sub category, and if they are similar to each other, they are classified into the same category according to their characteristics. Therefore, in this paper, we not only consider the purchase order of products to compare two purchase transaction sequences, but also calculate their similarity by giving a higher score if they are in the same category in spite of their difference. Especially, in order to choose the best similarity measure that directly affects the calculation performance of the purchase transaction sequences, we have compared the performance of three representative similarity measures, the Levenshtein distance, dynamic time warping distance, and the Needleman-Wunsch similarity. We have extended the existing methods to take into account the product taxonomy. For conventional similarity measures, the comparison of goods in two sequences is calculated by simply assigning a value of 0 or 1 according to whether or not the product is matched. However, the proposed method is subdivided to have a value between 0 and 1 using the product taxonomy tree to give a different degree of relevance between the two products, even if they are different products. Through experiments, we have confirmed that the proposed method was measured the similarity more accurately than the previous method. Furthermore, we have confirmed that dynamic time warping distance was the most suitable measure because it considered the degree of association of the product in the sequence and showed good performance for two sequences with different lengths.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.7
• /
• pp.685-691
• /
• 2020

In this study, an algorithm that recognizes numbers from 0 to 9 was developed using the data obtained after tracking hand movements using the echo signal of a millimeter-wave radar sensor at 77 GHz. The echo signals obtained from the radar sensor by detecting the motion of a hand gesture revealed a cluster of irregular dots due to the difference in scattering cross-sectional area. A valid center point was obtained from them by applying a K-Means algorithm using 3D coordinate values. In addition, the obtained center points were connected to produce a numeric image. The recognition rate was compared by inputting the obtained image and an image similar to human handwriting by applying the smoothing technique to a CNN (Convolutional Neural Network) model trained with MNIST (Modified National Institute of Standards and Technology database). The experiment was conducted in two ways. First, in the recognition experiments using images with and without smoothing, average recognition rates of 77.0% and 81.0% were obtained, respectively. In the experiment of the CNN model with augmentation of learning data, a recognition rate of 97.5% and 99.0% on average was obtained in the recognition experiment using the image with and without smoothing technique, respectively. This study can be applied to various non-contact recognition technologies using radar sensors.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.19 no.3
• /
• pp.153-163
• /
• 2017

The objective of this study is to enhance the model's speed of estimating weather variables (e.g., minimum/maximum temperature, sunshine hour, PRISM (Parameter-elevation Regression on Independent Slopes Model) based precipitation), which are applied to the Agrometeorological Early Warning System (http://www.agmet.kr). The current process of weather estimation is operated on high-performance multi-core CPUs that have 8 physical cores and 16 logical threads. Nonetheless, the server is not even dedicated to the handling of a single county, indicating that very high overhead is involved in calculating the 10 counties of the Seomjin River Basin. In order to reduce such overhead, several cache and parallelization techniques were used to measure the performance and to check the applicability. Results are as follows: (1) for simple calculations such as Growing Degree Days accumulation, the time required for Input and Output (I/O) is significantly greater than that for calculation, suggesting the need of a technique which reduces disk I/O bottlenecks; (2) when there are many I/O, it is advantageous to distribute them on several servers. However, each server must have a cache for input data so that it does not compete for the same resource; and (3) GPU-based parallel processing method is most suitable for models such as PRISM with large computation loads.

• The KIPS Transactions:PartD
• /
• v.18D no.1
• /
• pp.9-22
• /
• 2011

In this paper we present an efficient solution to rotation?invariant boundary image matching. Computing the rotation-invariant distance between image time-series is a time-consuming process since it requires a lot of Euclidean distance computations for all possible rotations. In this paper we propose a novel solution that significantly reduces the number of distance computations using the envelope-based lower bound. To this end, we first present how to construct a single envelope from a query sequence and how to obtain a lower bound of the rotation-invariant distance using the envelope. We then show that the single envelope-based lower bound can reduce a number of distance computations. This approach, however, may cause bad performance since it may incur a larger lower bound by considering all possible rotated sequences in a single envelope. To solve this problem, we present a concept of rotation interval, and using the rotation interval we generalize the envelope-based lower bound by exploiting multiple envelopes rather than a single envelope. We also propose equi-width and envelope minimization divisions as the method of determining rotation intervals in the multiple envelope approach. Experimental results show that our envelope-based solutions outperform existing solutions by one or two orders of magnitude.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.43 no.6
• /
• pp.507-515
• /
• 2015

This paper proposes a filtering method for effective state vector estimation of highly maneuvering target. It is needed to hit the point called 'sweet spot' to increase the kill probability in missile interception. In paper, a filtering method estimates the length of a moving target tracked by a frequency modulated continuous wave (FMCW) radar. High resolution range profiles (HRRPs) is generated from the radar echo signal and then it's integrated into proposed filtering method. To simulate the radar measurement which is close to real, the study on the properties of scattering point of the missile-like target has been conducted with ISAR image for different angle. Also, it is hard to track the target efficiently with existing Kalman filters which has fixed measurement noise covariance matrix R. Therefore the proposed method continuously updates the covariance matrix R with sensor measurements and tracks the target. Numerical simulations on the proposed method shows reliable results under reasonable assumptions on the missile interception scenario.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.10A
• /
• pp.1131-1138
• /
• 2004

In Hierarchical Mobile Ipv6 (HMIPv6) networks, the mobility anchor point (MAP) handles binding update (BU) procedures locally to reduce signaling overhead for mobility. However, as the number of mobile nodes (MNs) handled by the MAP increases, the MAP suffers from the overhead not only to handle signaling traffic but also to Process data tunneling traffic. Therefore, it is important to control the number of MNs serviced by the MAP, in order to mitigate the burden of the MAP. We propose an adaptive load control scheme, which consists of two sub-algorithms: threshold-based admission control algorithm and session-to-mobility ratio (SMR) based replacement algorithm. When the number of MNs at a MAP reaches to the full capacity, the MAP replaces an existing MN at the MAP, whose SMR is high, with an MN that just requests binding update. The replaced MN is redirected to its home agent. We analyze the proposed load control scheme using the .Markov chain model in terms of the new MN and the ongoing MN blocking probabilities. Numerical results indicate that the above probabilities are lowered significantly compared to the threshold-based admission control alone.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.1
• /
• pp.113-120
• /
• 2020

The multi-view video is a collection of multiple videos capturing the same scene at different viewpoints. Thus, there is an advantage of providing for user oriented view pointed video. This paper is suggested that the compression performance of layered depth image structure expression has improved by using more improved method. We confirm the data size of layer depth image by encoding H.264 technology and the each performances of reconstructed images. The H.264/AVC technology has easily extended for H.264 technology of video contents. In this paper, we suggested that layered depth structure can be applied for an efficient new image contents. We show that the huge data size of multi-view video image is decreased, and the higher performance of image is provided, and there is an advantage of for stressing error restoring.

• Journal of Korea Multimedia Society
• /
• v.9 no.8
• /
• pp.1010-1019
• /
• 2006

CAD system is a very important technology in designing the products which we are using today. This CAD technology has been enlarging its area into 3D CAD systems with the development of computer graphics technologies. In particular, such advances have been realized in special area such as the CAD system for designing shoes. However, because it is difficult to design a 3D CAD system, most CAD developments except for major CAD software companies usually have employed the AutoCAD which realizes 3D concepts. Also, many shoes' designers have been skilled with 2D shoe CAD systems and then they can make high achievements. Therefore, we design a new shoe CAD system with advantages of 2D CAD system and 3D CAD system. We also implement the scheme to transfer the data between 2D CAD system and 3D CAD system and also grading technologies to get patterns of various shoes size from a basic shoe's design.