• Journal of the Korean Society for information Management
• /
• v.20 no.3
• /
• pp.241-259
• /
• 2003

The Basic Semantic Register(BSR) is and official ISO register designed for interoperability among eBusiness and EDI systems. The entities registered in the current BSR are not defined in a machine-understandable way, which renders automatic extraction of structural and relationship information from the register impossible. The purpose of this study is to offer a framework for designing an ontology that can provide semantic interoperability among BSR-based systems by defining data structures and relationships with RDF and OWL, similar meaning by the 'equivalentClass' construct in OWL, the hierachical relationships among classes by the 'subClassOf' construct in RDF schema, definition of any entities in BSR by the 'label' construct in RDF schema, specification of usage guidelines by the 'comment' construct in RDF schema, assignment of classes to BSU's by the 'domain' construct in RDF schema, specification of data types of BSU's by the 'range' construct in RDF schema. Hierarchical relationships among properties in BSR can be expressed using the 'subPropertyOf' in RDF schema. Progress in semantic interoperability can be expected among BSR-based systems through applications of semantic web technology suggested in this study.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.7C
• /
• pp.652-657
• /
• 2007

Optical subcarrier-multiplexed (OSCM) labels in optical label switching networks have been detected using interleavers composed of fiber-optic Mach-Zehnder interferometer. 10-GHz optical single-/double-sideband signals generated from dual-electrode Mach-Zehnder intensity modulator have been used as the OSCM labels. In the case of single-sideband signals, the upper-sideband was observed to be suppressed about 16.8 dB compared with the lower-sideband from the optical spectrum measured at the label extraction output. For the case of double-sideband signals, both sidebands appeared with small insertion loss at the interleaver output. Since we used the phase-shift method to generate single-sideband signals, the power level of the single-sideband was higher by 3 dB than that of the double-sidebands.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.414-423
• /
• 2020

In this paper, we propose a Dilated Convolution Gate Linear Unit (DCGLU) to mitigate the lack of sparsity and small receptive field problems caused by the segmentation map extraction process in sound event detection with weak labels. In the advent of deep learning framework, segmentation map extraction approaches have shown improved performance in noisy environments. However, these methods are forced to maintain the size of the feature map to extract the segmentation map as the model would be constructed without a pooling operation. As a result, the performance of these methods is deteriorated with a lack of sparsity and a small receptive field. To mitigate these problems, we utilize GLU to control the flow of information and Dilated Convolutional Neural Networks (DCNNs) to increase the receptive field without additional learning parameters. For the performance evaluation, we employ a URBAN-SED and self-organized bird sound dataset. The relevant experiments show that our proposed DCGLU model outperforms over other baselines. In particular, our method is shown to exhibit robustness against nature sound noises with three Signal to Noise Ratio (SNR) levels (20 dB, 10 dB and 0 dB).

• Analytical Science and Technology
• /
• v.15 no.5
• /
• pp.475-481
• /
• 2002

The determination of phthalates and adipate in natural mineral water and its container is described. Phthalates and adipate were extracted from natural mineral water by liquid-liquid extraction with methylene chloride, concentrated and then injected in GC-MS (SIM). Phthalates and adipate from 1) PET, cap, label and glue were extracted in Soxhlet with 50 mL of carbon tetrachloride, purified with silicagel and detected with GC-MS (SIM). Peak shapes and quantitation of phthalates and adipate were excellent, with linear calibration curves over a range of $0.1{\sim}10{\mu}g/L$ in water sample ($r^2$ > 0.996) and over a range of $1{\sim}1,000{\mu}g/Kg$ in solid samples ($r^2$>0.994). The detection limits of analytes were $0.002{\sim}0.010{\mu}g/L$ in water and $0.01{\sim}0.02{\mu}g/Kg$ in solid samples. Five kinds of natural mineral water samples, two PETs, two labels, two caps and two glues were quantified by the described procedure. As a results, the concentrations of total phthalates in natural mineral water ranged from ND ~ 1.2 ng/mL. Otherwise, the concentrations of total phthalate extracted from PET ranged from 0.55 ~ 1.2 mg/Kg. We found that the accurate determination of phthalte and adipate in natural mineral water and container must be considered blank correction and the removal of label and glue in PET sample.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.10
• /
• pp.4968-4986
• /
• 2017

In this paper, we primarily address the difficulty of automatic generation of a plausible depth map from a single image in an unstructured environment. The aim is to extrapolate a depth map with a more correct, rich, and distinct depth order, which is both quantitatively accurate as well as visually pleasing. Our technique, which is fundamentally based on a preexisting DepthTransfer algorithm, transfers depth information at the level of superpixels. This occurs within a framework that replaces a pixel basis with one of instance-based learning. A vital superpixels feature enhancing matching precision is posterior incorporation of predictive semantic labels into the depth extraction procedure. Finally, a modified Cross Bilateral Filter is leveraged to augment the final depth field. For training and evaluation, experiments were conducted using the Make3D Range Image Dataset and vividly demonstrate that this depth estimation method outperforms state-of-the-art methods for the correlation coefficient metric, mean log10 error and root mean squared error, and achieves comparable performance for the average relative error metric in both efficacy and computational efficiency. This approach can be utilized to automatically convert 2D images into stereo for 3D visualization, producing anaglyph images that are visually superior in realism and simultaneously more immersive.

• Journal of Internet Computing and Services
• /
• v.6 no.3
• /
• pp.129-145
• /
• 2005

This paper proposes the web information extraction system that extracts the pre-defined information automatically from web documents (i.e, HTML documents) and integrates the extracted information, The system recognizes entities without lables by the probabilistic based entity recognition method and extends the existing domain knowledge semiautomatically by using the extracted data, Moreover, the system extracts the sub-linked information linked to the basic page and integrates the similar results extracted from heterogeneous sources, The experimental result shows that the system extracts the sub-linked information and uses the probabilistic based entity recognition enhances the precision significantly against the system using only the domain knowledge, Moreover, the presented system can the more various information precisely due to applying the system with flexibleness according to domains, Because bath the semiautomatic domain knowledge expansion and the probabilistic based entity recognition improve the quality of the information, the system can increase the degree of user satisfaction at its maximum. Thus, this system can satisfy the intellectual curiosity of users from movie sites, performance sites, and dining room sites, We can construct various comparison shopping mall and contribute the revitalization of e-business.

• Journal of Digital Convergence
• /
• v.17 no.11
• /
• pp.233-240
• /
• 2019

In order to predict and respond to cyber attacks, a number of security companies quickly identify the methods, types and characteristics of attack techniques and are publishing Security Intelligence Reports(SIRs) on them. However, the SIRs distributed by each company are huge and unstructured. In this paper, we propose a framework that uses five analytic techniques to formulate a report and extract key information in order to reduce the time required to extract information on large unstructured SIRs efficiently. Since the SIRs data do not have the correct answer label, we propose four analysis techniques, Keyword Extraction, Topic Modeling, Summarization, and Document Similarity, through Unsupervised Learning. Finally, has built the data to extract threat information from SIRs, analysis applies to the Named Entity Recognition (NER) technology to recognize the words belonging to the IP, Domain/URL, Hash, Malware and determine if the word belongs to which type We propose a framework that applies a total of five analysis techniques, including technology.

• Journal of Institute of Control, Robotics and Systems
• /
• v.3 no.5
• /
• pp.511-519
• /
• 1997

This paper presents the automatic recognition algorithm of the license number in on vehicle image. The proposed algorithm uses the correlation coefficient and Hough transform to detect license plate. The m/n ratio reduction is performed to save time and memory. By the correlation coefficient between the standard pattern and the target pattern, licence plate area is roughly extracted. On the extracted local area, preprocessing and binarization is performed. The Hough transform is applied to find the extract outline of the plate. If the detection fails, a smaller or a larger standard pattern is used to compute the correlation coefficient. Through this process, the license plate of different size can be extracted. Two algorithms to each separate number are proposed. One segments each number with projection-histogram, and the other segments each number with the label. After each character is separated, it is recognized by the neural network. This research overlomes the problems in conventional methods, such as the time requirement or failure in extraction of outlines which are due to the processing of the entire image, and by processing in real time, the practical application is possible.

• Food Science of Animal Resources
• /
• v.41 no.1
• /
• pp.110-121
• /
• 2021

Meat industries utilize plant material such as celery in cured meat products. Extraction of valuable bioactive compounds, nitrates and nitrites often involves processes that increase cost or lack sustainability. Thus, this study investigated the effect of ball-milled celery powders (CP) on the physicochemical, antioxidant, and antimicrobial properties along with curing efficiency in comminuted meat product. Pork sausages loaded with CPs with different average particle sizes: 265 ㎛ (T1), 68 ㎛ (T2) and 7 ㎛ (T3) were compared to those added without and with sodium nitrite (150 ppm). The a⁎ values were increased for sausages with larger particle size. The L⁎ values decreased for all CPs. Residual nitrite for all particle sizes increased in the earlier stages and decreased at the end of storage period. The curing efficiency also increased for larger size particles with an increase until day 9 followed by a gradual decrease. Superfine CP had a tendency to improve the antioxidant activities. The antimicrobial activity of CPs was not comparable with nitrite added sausages. The textural parameters remained unaffected by particle size. Thus, instead of extracts or juices, micronized CPs could be used to improve the antioxidant activities and curing efficiency of label friendly reformulated meat products.

• Korean Journal of Remote Sensing
• /
• v.37 no.4
• /
• pp.719-731
• /
• 2021

This study proposes a two-stage hybrid classification model for crop classification using multi-temporal remote sensing images; the model combines feature embedding by using an autoencoder (AE) with a convolutional neural network (CNN) classifier to fully utilize features including informative temporal and spatial signatures. Long short-term memory (LSTM)-based AE (LAE) is fine-tuned using class label information to extract latent features that contain less noise and useful temporal signatures. The CNN classifier is then applied to effectively account for the spatial characteristics of the extracted latent features. A crop classification experiment with multi-temporal unmanned aerial vehicle images is conducted to illustrate the potential application of the proposed hybrid model. The classification performance of the proposed model is compared with various combinations of conventional deep learning models (CNN, LSTM, and convolutional LSTM) and different inputs (original multi-temporal images and features from stacked AE). From the crop classification experiment, the best classification accuracy was achieved by the proposed model that utilized the latent features by fine-tuned LAE as input for the CNN classifier. The latent features that contain useful temporal signatures and are less noisy could increase the class separability between crops with similar spectral signatures, thereby leading to superior classification accuracy. The experimental results demonstrate the importance of effective feature extraction and the potential of the proposed classification model for crop classification using multi-temporal remote sensing images.