Earticle

문자 인코딩은 문자나 기호를 컴퓨터로 표현하기 위해 사용되는 방법이며 문자 인코딩 판별 소프트웨어들이 존재한다. 기존의 널리 쓰이는 인코딩 판별 소프트웨어인“uchardet”의 경우 변조되지 않은 일반 문서의 인코딩 판별 정확도는 91.39% 이지만 언어 판별 정확도는 32.09%에 불과하다. 또한 문서가 치환 암호에 의해 암호화 된 경우 인코딩 판별 정확도는 3.55%, 언어 판별 정확도는 0.06%로 매우 낮은 정확도를 보였다. 따라서 본 논문에서는Deep learning 알고리즘인 LSTM(Long Short-Term Memory)을 이용한 문서의 인코딩 및 언어 판별 방법을제안하며, 기존의 인코딩 판별 소프트웨어“uchardet”보다 뛰어난 결과를 보였다. 제안하는 방법을 이용한 일반 문서의 인코딩 판별 정확도는 99.89%이며, 언어 판별 정확도는 99.92%이다. 또한 문서가 치환 암호에 의해 암호화된 경우에는 제안하는 방법의 인코딩 판별 정확도는 99.26%이며, 언어 판별 정확도는 99.77%로 매우 뛰어나다.

Character encoding is the method used to represent characters or symbols on a computer, and there are many encoding detection software tools. For the widely used encoding detection software“uchardet”, the accuracy of encoding detection of unmodified normal text document is 91.39%, but the accuracy of language detection is only 32.09%. Also, if a text document is encrypted by substitution, the accuracy of encoding detection is 3.55% and the accuracy of language detection is 0.06%. Therefore, in this paper, we propose encoding and language detection of text document using the deep learning algorithm called LSTM(Long Short-Term Memory). The results of LSTM are better than encoding detection software“uchardet”. The accuracy of encoding detection of normal text document using the LSTM is 99.89% and the accuracy of language detection is 99.92%. Also, if a text document is encrypted by substitution, the accuracy of encoding detection is 99.26%, the accuracy of language detection is 99.77%.

인코딩(encoding)은 컴퓨터에서 사용되는 음성, 영상 및 텍스트 문자를 부호화하는 방법을 가리킨다. 그러므로 특정 데이터를 열람하기 위해서는 해당 인코딩 정보를 알아야하며, 데이터마다 인코딩을 판별해주는 알고리즘들이 존재한다. 하지만 실제 음원이나 문서를 송수신하는 과정에서 패킷 손실이 발생할 수 있으며, 특히 무선 통신망에서패킷 스니핑으로 정보를 가로챌 경우 손실률은 더욱 증가되어 인코딩 기법 판별에 어려움이 발생한다. 본 논문에서는 이러한 오류가 발생한 데이터의 인코딩 기법 판별율을 향상시키기 위해 기존의 문자 인코딩 기법 판별 프로그램인 'uchardet'에 Bit-shift 알고리즘을 적용하여 성능 향상을 이루었다. 알고리즘의 성능 평가를 위해 임의의 한글및 일본어 텍스트 파일에 손실률(loss rate)을 적용하여 부분적으로 데이터가 소실된 인코딩 파일을 생성하여 결과를 비교하였다. 그 결과, 패킷이 손실된 데이터에서 Bit-shift 알고리즘을 적용한 'uchardet-bitshift' 경우 기존의 알고리즘보다 더 나은 성능을 보였다. 한국어 인코딩의 경우 기존의 uchardet는 0.005% 손실률까지 100%의정확도를 보이고 1%보다 높은 손실률에서는 인코딩을 전혀 판별해 내지 못한데 비해, Bit-shift 알고리즘을 적용할 경우 0.05%의 손실률에도 100%의 정확도를 보였으며 그보다 큰 손실률에서도 해당 인코딩을 판별해냈다. 또한 한자어를 많이 포함하는 일본어의 경우 손실률이 높아질수록 중국어 인코딩으로 잘못 판별하는 경향을 보였다.시뮬레이션 분석 결과, Bit shift 알고리즘을 추가하여 기존 인코딩 기법 판별 알고리즘의 개선이 가능하였다.

In computer science, an encoding is a standardization of converting information to one format for audio, video or text. Therefore, the encoding information of the data should be known to open and read it and there are algorithms detecting encoder of the data. However, some informations of data could be disappeared by packet loss when transmitted on network, especially, if the data is snatched by packet sniffing or eavesdropping from wireless communications. In this paper, we improve the performance of encoding detection algorithm of 'uchardet' program for multi-byte encoded data with errors based on bit-shift algorithm. To simulate the performance, we generated Korean and Japanese text data with errors that is removed some random bits at random positions. Then the detection algorithm are tested using the data and 'uchardet-bitshift' showed better performance than 'uchardet'. When Korean texts are used, 'uchardet' could detect perfectly with ≤0.005% errors but it showed 0% detection rate with ≥1% errors while 'uchardet-bitshift' detected perfectly with ≤0.05% errors and it showed correct detection cases with ≥1% errors. Japanese texts with errors tend to report falsely as Chinese encoding because Japanese texts include lots of Chinese characters. As a results, we improved encoding detection algorithms by applying bit shift operation.

트랜스포머 모델의 핵심 요소인 토크나이저는 숫자 형태의 데이터를 제대로 이해하지 못한다. 따라서 패킷 페이로드 를 문장처럼 학습하여 실제 네트워크에서 동작 가능한 트랜스포머 기반의 침입탐지 모델을 구축하기 위해서는 16진수 형태의 패킷 페이로드를 문자 형태로 변환하는 것이 필요하다. 이러한 문제 인식 하에 본 연구에서는 3종의 문자 인코 딩 방식을 적용하여 패킷 페이로드를 숫자 및 문자 형태로 변환한 후 트랜스포머 모델에 학습시키면서 모델의 탐지성 능이 어떻게 달라지는지를 분석하였다. 성능 분석 실험을 위한 데이터세트는 UNSW-NB15 데이터세트에 포함된 PCAP 파일에서 패킷 페이로드를 추출하여 구성하였으며, 학습 모델은 RoBERTa를 사용하였다. 실험 결과, ISO-8859-1 인코딩이 이진분류 및 다중분류에서 가장 우수한 성능을 달성하는 것으로 확인되었으며, 토큰의 수를 512 개로 설정하고 최대 에포크를 15회로 증가한 경우에 다중분류 정확도가 88.77%까지 향상되었다.

A tokenizer, which is a key component of the Transformer model, lacks the ability to effectively comprehend nu merical data. Therefore, to develop a Transformer-based intrusion detection model that can operate within a real-w orld network environment by training packet payloads as sentences, it is necessary to convert the hexadecimal pack et payloads into a character-based format. In this study, we applied three character encoding methods to convert pa cket payloads into numeric or character format and analyzed how detection performance changes when training the m on transformer architecture. The experimental dataset was generated by extracting packet payloads from PCAP fi les included in the UNSW-NB15 dataset, and the RoBERTa was used as the training model. The experimental resul ts demonstrate that the ISO-8859-1 encoding scheme achieves the highest performance in both binary and multi-cla ss classification. In addition, when the number of tokens is set to 512 and the maximum number of epochs is set to 15, the multi-class classification accuracy is improved to 88.77%.

Since corpus is a pile of everyday language usage, using computing tools is essential in collecting, sifting, mining and using the meaningful data from the massive text data. In this paper, we introduce two tools for handling the large scale corpus; IMS Corpus Workbench (CWB) and Sketch Engine. The architecture of the tools is the inverted index model as a type of reference database, providing corpus handlers with speed and extendibility. The limit of CWB lies in the Western language character unicode system (ISO-8859), causing unsatisfactory handling of Korean in the full-fledged scale. We need to consider more suitable architectural design for searching, storing and user-friendly interface in case of large scale corpus in Korean.

본 논문은 원 데이터를 암호화하여 URL(Uniform Resource Locator)정보 형식으로 변형하고, URL정보를 QR코드로 인코딩함으로써 원 데이터의 저장없이 QR코드를 매개로 데이터를 기록하고 해독할 수 있도록 하는 스마 트한 QR코드에 의한 비저장식 데이터 기록 시스템 및 비저장식 데이터 제공방법을 제시하는 데 목적이 있다. 연구 의 방법은 선행기술연구와 문헌조사를 통해 연구의 차별성을 두기 위한 방법을 제시했으며 분석결과, 온라인 관리 서버에 구축되는 시스템으로서, 데이터 입력 신호별 매칭 암호가 저장된 암호DB, QR코드 생성 명령에 따라 입력 데이터를 암호DB에서 매칭되는 암호정보로 치환하여 관리서버 도메인네임의 하위 로케이터로 결합시킨 URL정보 로 변환시키고, URL정보를 QR코드로 인코딩하는 코드 생성부, 그리고 URL접속신호 수신에 따라, URL정보를 입력 데이터로 변환 제공하는 암호 해독부를 포함하는 QR코드에 의한 비저장식 데이터 기록시스템을 제공한다. 따라서 스마트한 QR방식의 데이터관리방식(기록과 제공)은 사용의 편리성과 공간의 제약이나 용량사용의 장애가 없다는 점을 밝히고 있다.

The purpose of this paper is to present a smart QR code recording system and a method of nonstorage data delivery that transforms the original data into a form of URL information by encrypting it and encoding the URL information into a QR code so that the QR code can be written and decrypted in a medium without storing the original data. The method of the study was presented by the prior art study and the literature research. Analysis results show that the system is built on the online administration server. The data input signal matching secret code is stored in DB, the QR code generation command converts input data from the password DB to the password information combined into the subordinate locator of the admin server's domain name, URL code. Therefore, the smart QR method of data management (recording and providing) indicates that there are no limitations in the ease and space of use or obstacles to capacity use.

양화사의 작용역 현상을 인간 언어 본연의 관점에서 바라보고 어떤 기제를 통해 일어나는 현상인지에 관한 심도 깊은 설명을 하고 기존의 연구에서의 한계점을 극복하고 개선하고자 본 연구에서는 양화사의 작용역의 중의성을 최소주의의 가장 새로운 방식인 Chomsky(2013, 2015)의 표찰 알고리즘(Labeling Algorithm) 접근법에서의 자질 공유(feature sharing)와 Obata와 Epstein 의 자질분리 병합(Feature-splitting Internal Merge)을 적극 수용하여 최소주의 접근법으로 재분석하고자 한다. 양화명사구는 양화자질[QU]을 가지고 있고 v*에 파이자질 이외 양화자질[QU]이 있다고 가정함으로써 이 통사체의 표찰은 자질공유를 통해 <QU, QU>로 결정된다고 제안하고자 한다. 이때 결정된 <QU, QU> 표찰은 부호화(encoding)되어 의미접합부(Conceptual-Intentional Interface ; CI Interface)에서 양화사의 중의적 해석을 가능하게 한다. 자유로운 병합이 전제된 최근의 최소주의 흐름속에서 Chomsky가 제안하는 표찰 알고리즘을 통해 양화사가 두 개인 문장에서 결정된 표찰들 간의 위계는 의미해석 접합부에서 의미정보를 전달하여 해석을 하는데 중요한 역할을 하고 있다고 전제하여 본 논문에서는 양화사 작용역 현상에 적용 가능한 세 가지 방법을 제안한다. 양화사가 두 개인 문장에서 첫째, 핵 v가 보유한 파이 자질과 양화 자질을 동시에 R로 상속하는 경우와 둘째, 두 자질 중 파이 자질을 먼저 R로 분리 병합하는 경우와 마지막으로 두 자질 중 양화 자질을 우선 R로 분리 병합하는 경우 세 가지 경우의 수가 발생한다. 이 세 가지 경우의 수 중에 첫 번째는 표찰 결정의 실패로 의미접합부에 무의미한 정보를 주지만, 두 번째 경우는 역(逆)의미 작용역을, 세 번째 경우는 표면 작용역의 해석적 정보를 주게된다. 다시 말해서, 양화사 의미의 중의성은 운용상의 부담(burden)을 주는 의무적 이동(obligatory movement)이 아니라 자유병합에 의한 결과물이며 이는 최소주의의 관점 내에서 작용역 현상의 간소화와 다양한 구문에 적용 가능한 통합된 설명을 시도하는데 그 의의가 있다.

A central goal of this paper is to present a new account of Quantifier scope ambiguity phenomena based on Chomsky's (2013, 2015) Labeling Algorithm (LA). In the account, this paper argues, without appealing to traditional movement operations akin to QR, that an inverse scope of QNP is attributed to the output of the essential operation in Narrow Syntax(NS) which is LA. Obata and Epstein's(2011) Feature-Splitting Internal Merge present empirical and theoretical advantages. Following their assumption, we suggest three viable options. First, R attracts both φ-feature and Quantifier feature on v* simultaneously. Secondly, R attracts Quantifier feature first. In the case that R attracts φ feature first is the last viable option. Given this view, we naturally hypothesize that φ-feature and Quantifier feature each move independently. In this sense, Feature splitting is the optimally exploiting independent mechanisms and enables us to capture derivationally different types of feature sharing. By deploying three viable options, we safely capture rigid scope and inverse scope. In summary, syntax of scope relies on three distinctive devices : Quantifier features, Feature-Splitting Internal Merge and Labeling Algorithm. Chomsky demonstrates that labels are needed for interface interpretation. In the spirit of minimalism, we elucidate scope ambiguity from a labeling perspective and this analysis sheds light on how determine a complicated syntactic object such as{XP[F],YP[F]} via labeling system and the quest for simplicity of Quantifier Scope Interaction.

본 연구는 재해현장에서 발생한 다양한 형태의 디지털 증거물을 채집・분석하는데 있어 디지털 포렌식 기법에 입각하여 효율적으로 재해 현장 디지털 데이터를 채증할 수 있는 e-Forensic 툴 개발에 관한 것이다. 현장에서 채집한 다양한 디지털 증거물이 법적 증거물로 채택되기 위해서는 디지털 증거물이 위변조가 없는 원본과 동일한 상태임을 입증해야만 하는데 이를 본 e-Forensic툴을 이용하여 검증하였다. 이 툴을 활용하면, 현장 수사 중 디지털 파일을 증거로 수집하는 경우에 현장에서 바로 해시값을 생성, 암호화된 디지털 파일에 부가함으로써 디지털 파일의 원본과 사본의 동일성을 보장하고 현장 증거의 무결성을 향상시킬 수 있는 효과가 있음을 증명하였다.

This research is about the development of e-forensic tool that extract & analyze different forms of digital evidence that individuals come across in a disaster scene. The tool utilizes digital forensic techniques which makes the tool efficient in any disaster analysis situation. In order for the forensic evidence to be selected as legal evidence, the evidence needs to be proven that it is in its original state with no forgery involved. This is where the e-forensic tool comes in, as its ability to collect digital evidence during investigation has proven; that the tool can keep the evidence in its original state and increase the integrity by generating hash TAG and adding the forensic evidence to a password encoded file.

기존의 바코드는 정보 배열이 나란히 나열된 선 모양을 가지며 이를 1차원 바코드라 부른다. 이에 반해 2차원 바코드는 점자방식 또는 모자이크방식 코드로 작은 정사각형 도는 직사각형 안에 정보를 표현한다. 2차원 바코드는 기존의 1차원 바코드보다 작은 공간에 많은 데이터를 표현 가능함으로써 보다 효율적인 바코드의 구현이 가능하다. 현재 ISO 국제 표준화된 2차원 바코드는 총 4가지로 분류되는데 QR Code, Data Matrix, PDF417, MaxiCode가 있다. 본 논문에서는 ISO 국제 표준화된 바코드 중 하나인 Data Matrix의 Base 256 모드에 대한 기본 개념, 구성 방법, 인코딩 및 디코딩 방법을 상세히 제안한다. Data Matrix 심벌에 저장된 데이터를 보다 효율 적으로 구성하기 위해 숫자, Alphanumeric 문자, 이진법에 따라 다른 인코딩, 디코딩 방법을 사용하게 되는데 본 논문에서는 이를 고려한 디코딩 방법에 초점을 맞춰 기술할 것이다.

Conventional bar code has the appearance of line bars and spaces, called as one-dimensional bar code. In contrast, the information in two-dimensional bar code is represented by either a small, rectangular or square with the types of mosaic and Braille. The two-dimensional bar code is much more efficient than one-dimensional bar code because it can allow to store and express large amounts of data in a small space and so far there is also a little information about decoding the Data Matrix in base 256 mode. According to the ISO international standards, there are four kinds of bar code: QR code, Data Matrix, PDF417, and Maxi code. In this paper, among them, we focus on describing the basic concepts of Data Matrix in base 256 mode, how to encode and decode them, and how to organize them in detail. In addition, Data Matrix can be organized efficiently depending on the modes of numeric, alphanumeric characters, and binary system and expecially, we focus on describing how to decode the Data Matrix code by four modes.

Relationship between reflection resonance and applied current density in Bragg photonic crystal has been investigated. Multiple bit encodes of distributed Bragg reflector features have been prepared by electrochemical etching of crystalline silicon by using various square wave current densities. Optical characterization of multi-encoding of distributed Bragg reflectors on porous silicon was achieved by Ocean optics 2000 spectrometer for the search of possible applications of multiple bit encoding of distributed Bragg reflectors such as multiplexed assays and chemical sensors. The morphology and cross-sectional structure of multi-encoded distributed Bragg reflectors was investigated by field emission scanning electron micrograph.

New method to encode multiple photonic features of Bragg type reflector on silicon wafer has been investigated. Multiple bit encodes of distributed Bragg reflector features have been prepared by electrochemical etching of crystalline silicon by using various square wave current densities. Optical characterization of multi-encoding of distributed Bragg reflectors on porous silicon was achieved by Ocean optics 2000 spectrometer for the search of possible applications of multiple bit encoding of distributed Bragg reflectors such as multiplexed assays and chemical sensors. The morphology and cross-sectional structure of multi-encoded distributed Bragg reflectors was investigated by field emission scanning electron micrograph.

This paper examines the Encoding Specificity Principle (ESP) and the related retrieval model (Synergistic Ecphoric Model of Retrieval) and assumption of the distinction between episodic and semantic memory. With the examination some difficulties with ESP and the related retrieval model and assumption are discussed.

Spiking Neural Networks (SNNs) offer promising solutions for efficient real-time processing of time-series data by closely emulating biological neuronal dynamics. However, existing encoding methods for converting raw input data into spike trains often introduce significant temporal distortions, complexity, or limitations in learnability, hindering their practical deployment. In this study, we propose the Filtered Temporal-Population (FTP) encoding method, a novel technique that integrates filtering operations into SNN encoding. FTP encoding effectively captures both temporal and spatial correlations within data segments while aligning inputs directly with the temporal axis, making it highly suitable for real-time applications. Evaluations on the MIT-BIH electrocardiogram dataset and other time-series datasets demonstrate that FTP encoding outperforms traditional encoding methods in terms of accuracy, speed, and robustness. Our findings highlight FTP encoding’s potential as a practical and effective solution for real-time SNN-based time-series classification tasks.

Cinnamyl alcohol dehydrogenase (CAD, EC 1.1.1.95), catalyzes the reduction of hydroxycinnamaldehydes to give hydroxycinnamyl alcohols, or "monolignols," the monomeric precursors of lignin. Lignins are important components of cell walls and lignified secondary cell walls play crucial roles in long distance transport of water and nutrients during plant growth and development and in plant defense against biotic and abiotic stresses. Here a cDNA clone containing a CAD gene, named as PgCAD, was isolated from a commercial medicinal plant Panax ginseng. PgCAD is predicted to encode a precursor protein of 177 amino acid residues, and its sequence shares high homology with a number of other plant CADS. The expression of PgCAD in adventitious roots and hairy roots of P. ginseng was analyzed using reverse transcriptase (RT)- PCR under various abiotic stresses such as salt, salicylic acid, wounding and chilling treatment that triggered a significant induction of PgCAD at different time points within 2-48 h post-treatment. This study revealed that PgCAD may help the plants to survive against various abiotic stresses.

Corynebacterium sp. SCR221107 was isolated from the rumen fluid of healthy male Holstein dairy cows from a research farm at Suncheon, Jeollanam-do, Korea. Corynebacterium sp. SCR221107 is a functional probiotic candidate that produces vitamin B12. All Corynebacterium sp. SCR221107 was sequenced using the PacBio RS II and Illumina HiSeq platforms and assembled de novo. The complete genome sequence of Corynebacterium sp. SCR221107 contained one circular chromosome (3,043,024 bp) with a guanine + cytosine (GC) content of 60.1%. Annotation analysis showed the presence of 2,639 protein-coding sequences, 15 rRNA genes, and 57 tRNA genes. Genome analysis found that Corynebacterium sp. SCR221107 encodes various genes associated with vitamin B12 synthesis and transport. The genomic information provided a detailed understanding of Corynebacterium sp. SCR221107, suggesting that this isolate may have potential probiotic applications.

By leveraging the characteristics of both quantum encoding and classical data channel, we present a unique and novel multiuser optical communication system that utilizes fascinating properties of the structured light for information transmission. Specifically, we propose an indoor orbital angular momentum (OAM) multiuser multiple-input multiple-output (MIMO) optical communication system, where the quantum encoding is performed and the multiuser transmission is then analyzed from the context of the classical data channel. In the proposed system, as the OAM beam propagates, its angular spread is analyzed. In addition, we show that the variance of the beam divergence increases with the increasing OAM index mode on the basis of the derived MIMO channel impulse response for OAM carrying Laguerre?Gaussian beams. Comparative studies reveal that the proposed system significantly outperforms the conventional non-orthogonal multiple access (NOMA) applied with MIMO.

The Lactococcus taiwanensis strain K_LL004 was isolated from the gut of a grasshopper (Oxya chinensis sinuosa) collected from local farm in Korea. L. taiwanensis strain K_LL004 is the functional probiotic candidate with an ability to hydrolyse plant polysaccharides. The complete genome of the L. taiwanensis strain K_LL004 contains one circular chromosome (1,995,099 bp) with a guanine + cytosine (GC) content of 38.8%. Moreover, 1,929 Protein- coding sequence, 19 rRNA genes, and 62 tRNA genes were identified based on results of annotation. L. taiwanensis strain K_LL004 has a gene, which encodes hydrolytic enzymes such as beta-glucosidase and beta-xylosidase, that hydrolyzes plant polysaccharides.

While motion estimation (ME) efficiently compresses video data by reducing temporal redundancy between frames, it does so at the cost of computational complexity. In particular, the complexity of sub-pixel ME (SME) is greatly limits fast video encoding because the SME requires a complexity-intensive interpolation process to generate imaginary pixels. Thus, this paper proposes a method that conditionally skips SME, including its interpolation process, by utilizing previously encoded motion information. Results show that the proposed method efficiently reduces both the total encoding time and memory access compared to the previous high efficiency video coding (HEVC) reference model.

Microsomal ?-3 fatty acid desaturase (FAD3) is an essential enzyme in the production of the n-3 polyunsaturated fatty acid α-linolonic acid during the seed developing stage. To understand the regulatory mechanism of the gene encoding the w-3 fatty acid desaturase, a genomic fragment corresponding to the previously isolated perilla seed PfFAD3 cDNA was amplified from perilla (Perilla frutescens Britt) by GenomeWalker PCR. Sequence analysis of the fragment provided with identification of a 1485-bp 5'- upstream region and a 241-bp intron in the open reading frame. To determine the tissue-specificity of the PfFAD3 gene expression, the 5'-upstream region was fused to the β-glucuronidase (GUS) gene and incorporated into Arabidopsis thaliana. Histochemical assay of the transgenic plants showed that GUS expression was restricted to seed and pollen, showing that PfFAD3 gene was exclusively expressed in those tissues.

Sentence voice does not change propositional structure but rearranges the proposition into a different information structure. The following experiment investigated previous research on the functional distinction in the encoding of active and passive sentences in terms of the given-new approach in relation to information structure. Furthemore, an attempt was made to test the hypothesis that the logical subject is the focus of the assertion in the passive and the active is presuppositionally neutral(Anisfeld and Klenbort, 1973; Klenbort and Anisfeld, 1974). The result of the experiment supported Klenbort and Anisfeld’s claim that the focal point of the information asserted by the sentence is the logical subject in the passive but failed to support their finding that the active is presuppositionally neutral. The test subjects’ responses indicate that the logical subject in both the active and the passive forms represents the assertional focus when the sentence is isolated from the context.

Recent deep learning models perform well in image and natural language processing. However, in tabular data, there is a problem that good performance is not achieved due to data-level problems. Recently, TabNet, a model that overcomes these shortcomings, has been widely used for tabular data learning. However, categorical variable data does not perform significantly in tabular data. To solve this problem, Catboost Encoding method is used to solve the problem. In the case of this model, the pre-processing of categorical variable data was well utilized to derive more performance than other models, and it showed better performance than other encoding techniques.