Earticle

본 연구에서는 상호 교체 가능한 일본어의 유사 표현에서 대해서 이들 표현의 교체 가능 성과 불가능성에 대한 인지적 이유를 제시하였다. 특히 일본어의 [かといって]와 [だからといっ て] 표현을 고찰 대상으로 하여 이들 표현의 교체 가능성과 불가능성에 대한 인지적 이유에 대해서 일본어 모국어 화자의 설문조사를 기반으로 한 네트워크 모델을 통해서 제시하였다. 구체적으로는 일본어 모국어 화자가 [だからといって]를 사용하는 경우는 [だからといって]의 앞 문장에 나오는 사건과 뒷 문장에 나오는 사건의 관계를 강하게 부정하는 경우로 뒤에 나타나 는 사건이 유일하다고 인지할 경우에 사용하는 한편, [かといって]의 사용은 [かといって]의 앞 문장에 나오는 사건과 뒷 문장에 나오는 사건의 관계를 부정할 때 사용되나 [だからといって] 와 다른 점은 뒤에 나타나는 사건이 여러 개 가운데 하나라고 인지할 경우에 사용된다는 사실을 제시하였다.

This study focused on Japanese expressions that can be used interchangeably and provides cognitive reasoning for the possibilities and impossibilities of substituting one of these expressions for another. In particular, the study focused on the Japanese expressions [dakaratoitte] and [katoitte], reflecting on substitution likelihood using a network model based on surveys by Japanese native speakers. It was found, for example, that Japanese native speakers use the expression [dakaratoitte], when the speaker strongly denies the relationship between the events that precede and follow [dakaratoitte ]; rather, they recognize the event as only being part of the sentence which is placed after [dakaratoitte]. In addition, although speakers use the expression [katoitte] to deny the relationship between the events from the sentences in front of [katoitte], [katoitte] is also used when the speaker recognizes that the event from the sentences after [katoitte] is one of the many events.

본고에서는 일반적으로 상호 대응하는 관계로 여겨지는 한국어와 일본어의 어휘를 대상으로 하여 이들 어휘가 가진 의미가 항상 상호 대응하는 관계로 나타나지 않는다는 사실에 착목하여 이들 양 어휘간의 대응가능성과 불가능성에 대한 인지언어학적 이유를 제시하고자 하였다. 구체적으로는 한국어 동사 ｢오르다(oreuda)｣와 이에 대응되는 일본어 동사 ｢上がる｣를 중심으로 이들이 의미적으로 서로 대응하는 경우와 대응하는 표현으로 나타나지 않는 경우의 예를 살펴 이들이 대응하여 나타나지 않는 경우를 중심으로 그 이유를 인지언어학적 관점에서 살펴보았다. 설명의 모델로서 이용한 것은 Langacker(2002)의 네트워크 모델(Network Model)이다. 네트워크 모델이란 어떠한 의미를 인지하는 데 있어서 처음에는 기존의 프로토타입과 마찬가지로 전형적인 범주에서부터 인지가 시작되기는 하나 그 인지적 범주가 확산되는 방향이 하나가 아닌 다각적인 방향이라는 점을 골자로 한 이론이다. 본고에서는 이와 같은 이론적 모델을 바탕으로 하여 한국어 동사 ｢오르다(oreuda)｣와 일본어 동사 ｢上がる｣의 의미 속성에 대해 조사하고 이들 동사의 인지적 범위를 네트워크 모델을 통해 도식화함으로써 이들이 각각 어떠한 전형적인 의미 속성을 바탕으로 어떠한 인지적 범위까지 확산된 의미 속성을 나타내는지 제시하였다. 또한 이를 통해 한일 양 언어에 있어 이들 두 동사가 서로 대응하는 표현으로 나타나지 않는 경우는 각 동사들의 인지적 범위의 차이가 다르기 때문이라는 사실을 확인하였다.

In the present paper, we focus on the meanings of Korean and Japanese vocabulary, which are generally considered to be in a mutually corresponding relationship, but do not always appear in a mutually corresponding relationship. It was intended to present an academic reason for this observation. Specifically, focusing on the Korean verb “oreuda” and the corresponding Japanese verb “agaru, we will examine examples where they semantically correspond to each other and where they do not appear in the corresponding expression. And these examples were analyzed from a cognitive linguistic point of view. The network model of Langacker (2002) was used as the explanatory model. The network model is a theory based on the point that, although recognition starts from a typical category like the existing prototype, the direction in which the cognitive category is spread is multifaceted. Based on such a theoretical model, this paper investigated the semantic properties of the Korean verb “oreuda” and the Japanese verb “agaru” and schematized the cognitive ranges of these verbs through a network model. Based on this, it was suggested to what extent the semantic attributes were spread out. In addition, it was also confirmed that sometimes these two verbs do not appear as corresponding expressions in both the Korean and Japanese languages and the reason ​​is due to the difference in the cognitive scope of each verb.

This study aimed to explore influencing factors on the establishment of the network system between public hospitals and to make a process evaluation of it. we analyzed the case of a strategic alliance contracted by a National University Hospital(NUH) and a Community Hospital(CH). Main points of the project were regular dispatch of clinical specialists in the NUH such as gastroenterologist and running teleradiology program. The NUH considered the improvement of it's image as a public hospital as a successful element of the network program. The provincial office which have to manage the CH satisfied with these program in terms of helping CH in need of clinical specialists. Staffs in the CH pointed out the problem of discontinuity for patients who visited the CH. Three institutes argued that continuous support of central government in the relevant institution and budget could play the most important role in the advance of the network system between public hospitals.

The most tedious and hectic job is to identify the tiny benthic animals by spending thousands of hour under the microscope, since all the fauna need to be counted, sorted, picked and permanently mounted on glass slides for taxonomic identification. All faunal identifications need a lot of preprocessing and it consumes a lot of time to identify a single specimen. Therefore, to reduce the complexity of many such procedures, combined with the desire to identify larger datasets, we came up with new software based on artificial intelligence which can automatically identify the benthic fauna through the microscopic images. In this paper, we propose a machine learning method for automatic visual identification through the images of the benthic fauna. To this end, we propose a neural network model, where we demonstrate that the proposed approach differentiates the fauna based on images. However, it works well with vast amounts of image data and significant computational resources.

In a duck cage, ducks are placed in various states. In particular, if a duck is overturned and falls or dies, it will adversely affect the growing environment. In order to prevent the foregoing, it was necessary to continuously manage the cage for duck growth. This study proposes a method using an object detection algorithm to improve the foregoing. Object detection refers to the work to perform classification and localization of all objects present in the image when an input image is given. To use an object detection algorithm in a duck cage, data to be used for learning should be made and the data should be augmented to secure enough data to learn from. In addition, the time required for object detection and the accuracy of object detection are important. The study collected, processed, and augmented image data for a total of two years in 2021 and 2022 from the duck cage. Based on the objects that must be detected, the data collected as such were divided at a ratio of 9 : 1, and learning and verification were performed. The final results were visually confirmed using images different from the images used for learning. The proposed method is expected to be used for minimizing human resources in the growing process in duck cages and making the duck cages into smart farms.

Clustering of vehicles is an important technique to reduce the high mobility effect of vehicles. This paper proposes an analytical model to evaluate the performance of a clustered vehicular ad hoc network (VANET). The analytical model is developed to evaluate three important parameters, namely packet delivery ratio, throughput, and delay. The results obtained from the analytical model are also accompanied by simulation results. This model can be further extended by researchers working on clustered VANET scenarios and will be helpful in modeling their protocols or algorithms. Furthermore, this model can verify the simulation results obtained from any network simulator.

Background: Social isolation reportedly induces excessive aggression and impairs hippocampal neurogenesis. In this study, we investigated whether Geijigadaehwang-tang (GDT), a traditional herbal formulation containing Rhei Radix et Rhizoma and Cinnamomi Ramulus, could ameliorate such behavioral and neurogenic deficits. Methods and Results: We assessed aggressive behaviors in socially isolated mice. Isolation mark- edly increased aggressive responses compared to sham controls. GDT treatment significantly reduced the number of attacks on days 3 and 7 and lowered both the attack duration and biting fre- quency. Immunohistochemical analysis of the dentate gyrus revealed that the isolation substantially reduced DCX-positive immature neuron numbers, whereas GDT administration partially restored these counts. Network pharmacology profiling has identified numerous GDT-associated com- pounds that interact with neuroendocrine and neurotransmission-related targets, including the dopaminergic, adrenergic, GABAergic, and CRH receptor pathways. Enrichment analysis high- lighted neuroactive ligand-receptor interaction, PI3K-Akt signaling, calcium signaling, apoptosis regulation, and steroid hormone biosynthesis as major functional clusters, supporting the role of GDT in neural plasticity and stress-related behavior. Conclusions: GDT attenuated isolation-induced aggression and partially rescued impaired hippo- campal neurogenesis. Integrated network analysis suggests that its multimodal actions might involve the modulation of neurotransmitter systems, neuroendocrine pathways, and pro-survival signaling cascades. GDT represents a promising multitarget therapeutic candidate for stress-related behavioral dysregulation and neurogenic impairment.

As more and more data is described, stored, exchanged and represented by XML, the abilities of information retrieval for XML document become increasingly important. However, the retrieval results to users are quite large. This paper gives a Bayesian network‐based model for ranking these large results. Each XML document is modeled through a Bayesian network, which can handle both structure and content for the document. And then this paper presents the inference for the probability of each document on the given query. Finally documents are ranked according to the probabilities in descent.

본 연구에서는 하도추적, 구체적으로는 상류단의 복수의 수위관측을 이용하여 하류단의 수위를 예 측하기 위하여 인공신경망 모델을 구성하였다. 대상하도는 금강유역의 용담댐과 대청댐 사이의 본 류이며, 상류단 입력자료로서 본류에 있는 수통, 호탄 관측소 관측수위와 지류인 송천 관측소 관측수 위를 고려하였다. 출력 값으로는 하류단의 옥천 관측소 수위를 3시간 및 6시간의 선행시간으로 예측 하도록 인공신경망 모형을 구성하였다. 인공신경망 모형의 구성은 단일 히든레이어를 중심으로 이 루어졌으며, Epoch number, Mini-batch size, Learning rate의 3가지 학습변수의 예측정확도에 대한 민 감도를 분석하였다. 인공신경망의 학습(Training), 시험(testing), 검증(validation)을 위해 2000년부터 2012년까지 13년간의 시수위자료를 이용하여 학습을 진행하였으며, 2013년부터 2014년의 2년간의 수위자료를 이용한 시험을 통해 최적의 모형을 선정하였다. 또한 선정된 최적의 모형을 이용하여 2015년부터 2016년까지의 수위예측을 수행하였다.

In this study, artificial neural network model was constructed in order to predict the water level. The target stream is the main stream between Yongdam Dam and Daecheong Dam in the Geum river Basin, and as input data at the upstream, the observation water level at the Sutong and the Hotan observatory and the water level at the Songcheon observatory, which is a tributary, are considered as input data. As an output value, an artificial neural network model was constructed to predict the water level of the Okcheon station at the downstream with 3 hours and 6 hours lead time. The artificial neural network model was constructed around a single hidden layer, and the sensitivity to the prediction accuracy of three learning variables was analyzed: Epoch number, Batch size, and Learning rate. For training, testing, and validation of artificial neural networks, learning was conducted using observed water level data for 13 years from 2000 to 2012, and water level data for two years from 2013 to 2014. The optimal model was selected through the used test. In addition, water level prediction from 2015 to 2016 was performed using the selected optimal model.

A stable neural network control scheme for unknown non-linear systems is developed in this paper. While the control variable is optimised to minimize the performance index, convergence of the index is guaranteed asymptotically stable by a Lyapnov control law. The optimization is achieved using a gradient descent searching algorithm and is consequently slow. A fast convergence algorithm using an adaptive learning rate is employed to speed up the convergence. Application of the stable control to a single input single output (SISO) non-linear system is simulated. The satisfactory control performance is obtained.

기존에 공공 데이터베이스 정보를 이용하여 자연재해로 인한 피해액을 추정하기 위한 회귀분석 모 형 개발 연구가 진행되었다. 그러나 이모형은 종속변수의 로그변환을 통해 독립변수와 종속변수 간의 비선형성을 반영했음에도 불구하고, 독립변수와 종속변수 간의 선형관계를 가정하는 회귀분석 모형의 고유한 특성으로 인해 예측 정확도 향상에 한계점을 나타냈다. 이에 본 연구에서는 독립변수 와 종속변수 간의 선형성 가정이 요구되지 않는 인공신경망 모형을 기반으로 자연재해 피해액을 좀 더 정확하게 예측할 수 있는 모형을 제시한다. 제안된 모형의 유효성을 검증하기 위해, 인공신경 망 모형을 이용한 피해액 예측치, 회귀분석 모형을 이용한 피해액 예측치, 자연재해 위험지표 평가결 과, 지역안전도 등급, 자연재해로 인한 실제 피해액을 비교⋅분석하였다. 분석 결과, 인공신경망을 이용하여 도출된 추정치가 기존의 회귀분석을 이용하여 도출된 추정치뿐만 아니라 자연재해 위험지 표와 지역안전도 평가결과에 비해서도 실제 자연재해 피해액과 더 높은 상관관계, 즉, 더 높은 예측 력을 나타내는 것을 확인할 수 있었다.

In the previous research, we developed a regression model for estimating damage of natural disasters based on the public database. Although this model considers nonlinearities among variables by using log transformation for the dependent variable, it reveals limitations in improving estimation accuracy because of its inherent characteristics of linearity assumption between independent and dependent variables. In this study, we proposed an artificial neural network (ANN) based model to predict the amount of damages due to natural disasters more accurately, which does not require the linearity assumption among the variables. For verification of the proposed model, we compared the model estimates with those from the regression model, including the Natural Disaster Risk Index (NDRI), Regional Safety Grades (RSG), and actual damage amounts. According to the results of analysis, we can confirm that the estimates from the ANN-based model reveal a higher correlation with the actual damage amounts than those from the regression model or the assessment results of NDRI and RSG.

본 연구에서는 수도권 4개 권역의 PM2.5 예보를 위해 심층신경망 모델을 개발하고, 계층별 관련성 전파를 사용하여 예 측결과에 대한 입력인자들의 기여도를 분석했다. 제안한 심층신경망 예보모델은 평균적으로 지수적중률 71%, 고농도 감 지 확률 65%, 오경보율 40% 정도의 성능을 보였다. 이러한 심층신경망 예보결과에 대한 입력인자들의 중요도를 분석하 기 위해 데이터를 지역 및 PM2.5 농도로 분류하여 계층별 관련성 전파를 수행했다. 그 결과 농도 및 지역에 상관없이 중 요도가 높은 인자는 압력, 온도, 이슬점온도, 상대습도, U, V, 이산화질소, 일산화탄소로 확인된 반면에 누적강수량, 아황 산가스, PM10, 오존은 중요도가 낮은 것으로 관찰되었다. 고농도 사례에서는 일사와 오존의 중요도가 낮아지고, 예측시 간에 근접한 시간대의 PM2.5의 중요도가 높아졌으며, CMAQ(The Community Multiscale Air Quality) 예측인자들의 중 요도도 다소 올라갔다. 지역별 입력인자들의 중요도는 대체로 유사했는데 인천과 경기북부의 경우 일사의 중요도가 낮아 지는 등 고농도 패턴의 인자중요도와 일부 비슷한 결과를 보였다. 이것은 인천과 경기북부가 다른 지역보다 고농도 데이 터가 더 자주 발생하기 때문으로 분석된다. 이러한 인자중요도 결과는 향후 고농도 적중률 향상 및 지역별 특성에 적합 한 예보모델을 개발하는데 활용하면 효과적일 것으로 기대한다.

In this study, we propose a deep neural network model for PM2.5 prediction in the four metropolitan areas, and analyze the contribution of input factors to the prediction results using layer-wise relevance propagation. The proposed deep neural network forecasting model exhibited an average exponential accuracy of 71%, a high-concentration detection probability of 65%, and a false alarm ratio of 40%. In order to analyze the importance of input factors for these deep neural network forecasting results, we classify the data into regional and PM2.5 concentration level and performed layer-wise relevance propagation. As a result, factors such as pressure, temperature, dew point temperature, relative humidity, U, V, nitrogen dioxide and carbon monoxide were found to be important, regardless of any region and concentration level. On the other hand, accumulated precipitation, sulfur dioxide, PM10, and ozone are shown to have lower importance. In highconcentration cases, the importance of solar radiation and ozone decreased, PM2.5 during time periods close to the prediction time increased, and the importance of CMAQ(The Community Multiscale Air Quality) predictive factors also showed some degree of elevation. The importance of input factors by region are generally similar; however, in the case of Incheon and Northern Gyeonggi Province, the importance of the radiation decreases, showing some similarities to the importance of factors in high concentration patterns. This is analyzed as being due to a higher occurrence of highconcentration data in Incheon and Northern Gyeonggi Province compared to other regions. These analyzing results are expected to be effective in developing forecasting models that improve the prediction accuracy of high concentration and match regional characteristics in the future.

A Collaborative Filtering-based recommender system collects all of customers’ transaction data and determines relevant customer set, called neighborhood, which are determined by representative measures such as Cosine or Pearson-correlation and then generates product recommendation lists from the transactions of neighborhood. Though such a procedure has been known to a very effective method, its computational overhead can be prohibitive when the customer base is large. At the same time, when the sparse level of the transaction data is high, it brings deterioration in the quality of recommendations. The paper proposes the use of a customer network, for recommendations that accommodate the large-scale and sparsity nature of the transaction dataset. What is proposed in this study is a more active form of social network application, recommender network, utilizing the fast diffusion and information sharing capability of social network. From the literature of bipartite graph, we formulate CF-based recommendation task as a network problem, and then we propose a microscopic process governing the link strength of dynamic networks as a recommendation process. In order to validate the effectiveness and the efficiency of the proposed method, we build a recommendation network for the product recommendation using real product transaction data and compare it against the traditional system based on collaborative filtering. Experiment results show that the microscopic process of the recommendation network is computationally more efficient than, but as accurate as, the global optimization process of the traditional recommender system.

The previous works in sensor networks security have focused on the aspect of confidentiality, authentication and integrity based on cryptographic primitives. There has been no prior work to assess the survivability in systematic way. Accordingly, this paper presents a survivability model of wireless sensor networks using software rejuvenation for dual adaptive cluster head. The survivability model has state transition to reflect status of real wireless sensor networks. In this paper, we only focus on a survivability model which is capable of describing cluster head compromise in the networks and able to switch over the redundant cluster head in order to increase the survivability of that cluster. Second, this paper presents how to enhance the survivability of sensor networks using software rejuvenation methodology for dual cluster head in wireless sensor network. We model and analyze each cluster as a stochastic process based on Semi Markov Process (SMP) and Discrete Time Markov Chain (DTMC). The proof of example scenarios and numerical analysis shows the feasibility of our approach.

In this paper, separated and optimized pattern database model is proposed. In order to improve efficiency of Network-based IDS, pattern database is classified by proper basis. Classification basis is decided by the specific Intrusions validity on specific target. Using this model, IDS searches only valid patterns in pattern database on each captured packets. In result, IDS can reduce system resources for searching pattern database. So, IDS can analyze more packets on the network. In this paper, proper classification basis is proposed and pattern database classified by that basis is formed. And its performance is verified by experimental results.