• 한국지하수토양환경학회지:지하수토양환경
• /
• 제23권2호
• /
• pp.1-14
• /
• 2018

The recent increase in demand for lithium has led to the development of new brine prospects, The brines are hosted in closed salar basin aquifers of two types that are mature halite salars and immature clastic salars. Salar brines also contain other elements of commercial interest, most notably potassium and boron. As a result, there has been a plethora of new exploration projects focused on the brines hosted in the aquifers of the intermontane-closed basins. The estimate of lithium resources and reserves in these salars depends on a detailed knowledge of aquifer geometry, porosity, and brine grade. Because the resource is in a fluid state, it has the propensity to move, mix, rearrange itself relatively rapidly during the course of a project lifetime, and lower recovery factors compared with most metalliferous and industrial mineral deposits due to reliance on pumping of the brine from wells for extraction. This is unlike any other type of metallic mineral resource and hence a different approach specially focusing on hydrogeology and brine hydrology is required for these prospects.

• Journal of Information Science Theory and Practice
• /
• 제9권2호
• /
• pp.66-82
• /
• 2021

Recommender Systems have gained immense popularity due to their capability of dealing with a massive amount of information in various domains. They are considered information filtering systems that make predictions or recommendations to users based on their interests and preferences. The more recent technology, Linked Open Data (LOD), has been introduced, and a vast amount of Resource Description Framework data have been published in freely accessible datasets. These datasets are connected to form the so-called LOD cloud. The need for semantic data representation has been identified as one of the next challenges in Recommender Systems. In a LOD-enabled recommendation framework where domain awareness plays a key role, the semantic information provided in the LOD can be exploited. However, dealing with a big chunk of the data from the LOD cloud and its integration with any domain datasets remains a challenge due to various issues, such as resource constraints and broken links. This paper presents the challenges of interconnecting and extracting the DBpedia data with the MovieLens 1 Million dataset. This study demonstrates how LOD can be a vital yet rich source of content knowledge that helps recommender systems address the issues of data sparsity and insufficient content analysis. Based on the challenges, we proposed a few alternatives and solutions to some of the challenges.

• Journal of Forest and Environmental Science
• /
• 제37권2호
• /
• pp.116-127
• /
• 2021

Predicting carbon distribution of soil aggregates is difficult due to complexity in organo-mineral formation. This limits global warming mitigation through soil carbon sequestration. Therefore, knowledge of land use effect on carbon stabilization requires quantification of soil mineral cations. The study was conducted to quantify carbon and base cations on soil mineral fractions in Natural Forest, Plantation Forest and Farm Land. Five 0.09 ha were demarcated alternately along 500 m long transect with an interval of 50 m in Natural Forest (NF), Plantation Forest (PF) and Farm Land (FL). Soil samples were collected with soil cores at 0-15, 15-30 and 30-45 cm depths in each plot. Soil core samples were oven-dried at 105℃ and soil bulk densities were computed. Sample (100 g) of each soil core was separated into >2.0, 2.0-1.0, 1.0-0.5, 0.5-0.05 and <0.05 mm aggregates using dry sieve procedure and proportion determined. Carbon concentration of soil aggregates was determined using Loss-on-ignition method. Mineral fractions of soil depths were obtained using dispersion, sequential extraction and sedimentation methods of composite soil samples and sieved into <0.05 and >0.05 mm fractions. Cation exchange capacity of two mineral fractions was measured using spectrophotometry method. Data collected were analysed using descriptive and ANOVA at α_0.05. Silt and sand particle size decreased while clay increased with increase in soil depth in NF and PF. Subsoil depth contained highest carbon stock in the PF. Carbon concentration increased with decrease in aggregate size in soil depths of NF and FL. Micro- (1-0.5, 0.5-0.05 and <0.05 mm) and macro-aggregates (>2.0 and 2-1.0 mm) were saturated with soil carbon in NF and FL, respectively. Cation exchange capacity of <0.05 mm was higher than >0.05 mm in soil depths of PF and FL. Fine silt (<0.05 mm) determine the cation exchange capacity in soil depths. Land use and mineral size influence the carbon and cation exchange capacity of Gambari Forest Reserve.

• 정보보호학회논문지
• /
• 제30권6호
• /
• pp.1053-1065
• /
• 2020

최근 하드웨어의 성능과 인공지능 기술이 발달함에 따라 육안으로 구분하기 어려운 정교한 가짜 동영상들이 증가하고 있다. 인공지능을 이용한 얼굴 합성 기술을 딥페이크라고 하며 약간의 프로그래밍 능력과 딥러닝 지식만 있다면 누구든지 딥페이크를 이용하여 정교한 가짜 동영상을 제작할 수 있다. 이에 무분별한 가짜 동영상이 크게 증가하였으며 이는 개인 정보 침해, 가짜 뉴스, 사기 등에 문제로 이어질 수 있다. 따라서 사람의 눈으로도 진위를 가릴 수 없는 가짜 동영상을 탐지할 수 있는 방안이 필요하다. 이에 본 논문에서는 Bidirectional Convolutional LSTM과 어텐션 모듈(Attention module)을 적용한 딥페이크 탐지 모델을 제안한다. 본 논문에서 제안하는 모델은 어텐션 모듈과 신경곱 합성망 모델을 같이 사용되어 각 프레임의 특징을 추출하고 기존의 제안되어왔던 시간의 순방향만을 고려하는 LSTM과 달리 시간의 역방향도 고려하여 학습한다. 어텐션 모듈은 합성곱 신경망 모델과 같이 사용되어 각 프레임의 특징 추출에 이용한다. 실험을 통해 본 논문에서 제안하는 모델은 93.5%의 정확도를 갖고 기존 연구의 결과보다 AUC가 최대 50% 가량 높음을 보였다.

• The Plant Pathology Journal
• /
• 제37권6호
• /
• pp.521-532
• /
• 2021

Knowledge and better understanding of functions of the microbial community are pivotal for crop management. This study was conducted to study bacterial structures including Acidovorax species community structures and diversity from the watermelon cultivated soils in different regions of South Korea. In this study, soil samples were collected from watermelon cultivation areas from various places of South Korea and microbiome analysis was performed to analyze bacterial communities including Acidovorax species community. Next generation sequencing (NGS) was performed by extracting genomic DNA from 92 soil samples from 8 different provinces using a fast genomic DNA extraction kit. NGS data analysis results revealed that, total, 39,367 operational taxonomic unit (OTU), were obtained. NGS data results revealed that, most dominant phylum in all the soil samples was Proteobacteria (37.3%). In addition, most abundant genus was Acidobacterium (1.8%) in all the samples. In order to analyze species diversity among the collected soil samples, OTUs, community diversity, and Shannon index were measured. Shannon (9.297) and inverse Simpson (0.996) were found to have the highest diversity scores in the greenhouse soil sample of Gyeonggi-do province (GG4). Results from NGS sequencing suggest that, most of the soil samples consists of similar trend of bacterial community and diversity. Environmental factors play a key role in shaping the bacterial community and diversity. In order to address this statement, further correlation analysis between soil physical and chemical parameters with dominant bacterial community will be carried out to observe their interactions.

• 우주기술과 응용
• /
• 제1권3호
• /
• pp.356-363
• /
• 2021

중적외선을 이용하여 획득한 영상은 상온 부근의 온도 측정에는 적당하지 않다는 것이 교과서적 상식이었으나, 최근 중적외선 센서를 이용한 위성 영상을 살펴보면 중적외선 센서를 이용하여 측정한 결과물도 상온 부근의 온도를 측정할 수 있다는 가능성을 보여주고 있다. 본 논문에서는 중적외선 센서를 가지는 위성영상의 상온 온도 정보 추출 가능성 및 정확도에 대해 살펴보고자 한다. 논문에서 검토된 중적외선 위성영상은 상온 부분의 온도를 잘 표시하였으며, 측정된 온도의 절대값으로서의 신뢰성에 관해서는, 태양광의 지표면 직접 반사에 의한 열전달량 영향과 대기에서 흡수되는 적외선 대기 흡수량의 영향은 비교적 작거나 일정한 값으로 볼 수 있지만, 비접촉식 온도계의 한계인 물성치의 의한 복사 계수(emissivity)의 불확실성 문제는 여전히 해결해야할 문제로 남아 있게 되었다.

• Imaging Science in Dentistry
• /
• 제52권3호
• /
• pp.239-244
• /
• 2022

Purpose: Despite the proliferation of numerous morphometric and anthropometric methods for sex identification based on linear, angular, and regional measurements of various parts of the body, these methods are subject to error due to the observer's knowledge and expertise. This study aimed to explore the possibility of automated sex determination using convolutional neural networks(CNNs) based on lateral cephalometric radiographs. Materials and Methods: Lateral cephalometric radiographs of 1,476 Iranian subjects (794 women and 682 men) from 18 to 49 years of age were included. Lateral cephalometric radiographs were considered as a network input and output layer including 2 classes(male and female). Eighty percent of the data was used as a training set and the rest as a test set. Hyperparameter tuning of each network was done after preprocessing and data augmentation steps. The predictive performance of different architectures (DenseNet, ResNet, and VGG) was evaluated based on their accuracy in test sets. Results: The CNN based on the DenseNet121 architecture, with an overall accuracy of 90%, had the best predictive power in sex determination. The prediction accuracy of this model was almost equal for men and women. Furthermore, with all architectures, the use of transfer learning improved predictive performance. Conclusion: The results confirmed that a CNN could predict a person's sex with high accuracy. This prediction was independent of human bias because feature extraction was done automatically. However, for more accurate sex determination on a wider scale, further studies with larger sample sizes are desirable.

• International Journal of Computer Science & Network Security
• /
• 제23권1호
• /
• pp.147-152
• /
• 2023

Mostly in motor fault detection the instantaneous values 3 axis vibration and 3phase current in time domain are acquired and converted to frequency domain. Vibrations are more useful in diagnosing the mechanical faults and motor current has remained more useful in electrical fault diagnosis. With having some experience and knowledge on the behavior of acquired data the electrical and mechanical faults are diagnosed through signal processing techniques or combine machine learning and signal processing techniques. In this paper, a single-layer LSTM based condition monitoring system is proposed in which the instantaneous values of three phased motor current are firstly acquired in simulated motor in in health and supply imbalance conditions in each of three stator currents. The acquired three phase current in time domain is then used to train a LSTM network, which can identify the type of fault in electrical supply of motor and phase in which the fault has occurred. Experimental results shows that the proposed single layer LSTM algorithm can identify the electrical supply faults and phase of fault with an average accuracy of 88% based on the three phase stator current as raw data without any processing or feature extraction.

• Journal of the Korean Association of Oral and Maxillofacial Surgeons
• /
• 제50권3호
• /
• pp.123-133
• /
• 2024

Dentoalveolar (DA) trauma, which can involve tooth, alveolar bone, and surrounding soft tissues, is a significant dentofacial emergency. In emergency settings, physicians might lack comprehensive knowledge of timely procedures, causing delays for specialist referral. This systematic review assesses the literature on isolated DA fractures, emphasizing intervention timing and splinting techniques and duration in both children and adults. This systematic review adhered to PRISMA guidelines and involved a thorough search across PubMed, Google Scholar, Semantic Scholar, and the Cochrane Library from January 1980 to December 2022. Inclusion and exclusion criteria guided study selection, with data extraction and analysis centered on demographics, etiology, injury site, diagnostics, treatment timelines, and outcomes in pediatric (2-12 years) and adult (>12 years) populations. This review analyzed 26 studies, categorized by age into pediatrics (2-12 years) and adults (>12 years). Falls were a common etiology, primarily affecting the anterior maxilla. Immediate management involved replantation, repositioning, and splinting within 24 hours (pediatric) or 48 hours (adult). Composite resin-bonded splints were common. Endodontic treatment was done within a timeframe of 3 days to 12 weeks for children and 2-12 weeks for adults. Tailored management based on patient age, tooth development stage, time elapsed, and resource availability is essential.

• 한국컴퓨터정보학회논문지
• /
• 제29권1호
• /
• pp.11-19
• /
• 2024

멀티에이전트는 전장 교전 상황, 무인 운송 차량 등 다양한 실제 협동 환경에 사용될 수 있다. 전장 교전 상황에서는 도메인 정보의 제한으로 즉각적인 보상(Dense Reward) 설계의 어려움이 있어 명백한 희소 보상(Sparse Reward)으로 학습되는 상황을 고려해야 한다. 본 논문에서는 전장 교전 상황에서의 아군 에이전트 간 협업 가능성을 확인하며, 희소 보상 환경인 Multi-Robot Warehouse Environment(RWARE)를 활용하여 유사한 문제와 평가 기준을 정의하고, 강화학습 라이브러리인 Ray RLlib의 QMIX 알고리즘을 사용하여 학습 환경을 구성한다. 정의한 문제에 대해 QMIX의 Agent Network를 개선하고 Random Network Distillation(RND)을 적용한다. 이를 통해 에이전트의 부분 관측값에 대한 패턴과 시간 특징을 추출하고, 에이전트의 내적 보상(Intrinsic Reward)을 통해 희소 보상 경험 획득 개선이 가능함을 실험을 통해 확인한다.