• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.12
• /
• pp.291-306
• /
• 2020

Nowadays, Data-Network-AI (DNA)-based intelligent services and applications have become a reality to provide a new dimension of services that improve the quality of life and productivity of businesses. Artificial intelligence (AI) can enhance the value of IoT data (data collected by IoT devices). The internet of things (IoT) promotes the learning and intelligence capability of AI. To extract insights from massive volume IoT data in real-time using deep learning, processing capability needs to happen in the IoT end devices where data is generated. However, deep learning requires a significant number of computational resources that may not be available at the IoT end devices. Such problems have been addressed by transporting bulks of data from the IoT end devices to the cloud datacenters for processing. But transferring IoT big data to the cloud incurs prohibitively high transmission delay and privacy issues which are a major concern. Edge computing, where distributed computing nodes are placed close to the IoT end devices, is a viable solution to meet the high computation and low-latency requirements and to preserve the privacy of users. This paper provides a comprehensive review of the current state of leveraging deep learning within edge computing to unleash the potential of IoT big data generated from IoT end devices. We believe that the revision will have a contribution to the development of DNA-based intelligent services and applications. It describes the different distributed training and inference architectures of deep learning models across multiple nodes of the edge computing platform. It also provides the different privacy-preserving approaches of deep learning on the edge computing environment and the various application domains where deep learning on the network edge can be useful. Finally, it discusses open issues and challenges leveraging deep learning within edge computing.

• Journal of Korea Water Resources Association
• /
• v.54 no.11
• /
• pp.969-983
• /
• 2021

Evapotranspiration, one of the hydrometeorological components, is considered an important variable for water resource planning and management and is primarily used as input data for hydrological models such as water balance models. The FAO56 PM method has been recommended as a standard approach to estimate the reference evapotranspiration with relatively high accuracy. However, the FAO56 PM method is often challenging to apply because it requires considerable hydrometeorological variables. In this perspective, the Hargreaves equation has been widely adopted to estimate the reference evapotranspiration. In this study, a set of parameters of the Hargreaves equation was calibrated with relatively long-term data within a Bayesian framework. Statistical index (CC, RMSE, IoA) is used to validate the model. RMSE for monthly results reduced from 7.94 ~ 24.91 mm/month to 7.94 ~ 24.91 mm/month for the validation period. The results confirmed that the accuracy was significantly improved compared to the existing Hargreaves equation. Further, the evaporative demand drought index (EDDI) based on the evaporative demand (E₀) was proposed. To confirm the effectiveness of the EDDI, this study evaluated the estimated EDDI for the recent drought events from 2014 to 2015 and 2018, along with precipitation and SPI. As a result of the evaluation of the Han-river watershed in 2018, the weekly EDDI increased to more than 2 and it was confirmed that EDDI more effectively detects the onset of drought caused by heatwaves. EDDI can be used as a drought index, particularly for heatwave-driven flash drought monitoring and along with SPI.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.1243-1244
• /
• 2022

In recent years, the growing interest in off-site construction has led to factories scaling up their manufacturing and production processes in the construction sector. Consequently, continuous large-scale site monitoring in low-variability environments, such as prefabricated components production plants (precast concrete production), has gained increasing importance. Although many studies on computer vision-based site monitoring have been conducted, challenges for deploying this technology for large-scale field applications still remain. One of the issues is collecting and transmitting vast amounts of video data. Continuous site monitoring systems are based on real-time video data collection and analysis, which requires excessive computational resources and network traffic. In addition, it is difficult to integrate various object information with different sizes and scales into a single scene. Various sizes and types of objects (e.g., workers, heavy equipment, and materials) exist in a plant production environment, and these objects should be detected simultaneously for effective site monitoring. However, with the existing object detection algorithms, it is difficult to simultaneously detect objects with significant differences in size because collecting and training massive amounts of object image data with various scales is necessary. This study thus developed a large-scale site monitoring system using edge computing and a small-object detection system to solve these problems. Edge computing is a distributed information technology architecture wherein the image or video data is processed near the originating source, not on a centralized server or cloud. By inferring information from the AI computing module equipped with CCTVs and communicating only the processed information with the server, it is possible to reduce excessive network traffic. Small-object detection is an innovative method to detect different-sized objects by cropping the raw image and setting the appropriate number of rows and columns for image splitting based on the target object size. This enables the detection of small objects from cropped and magnified images. The detected small objects can then be expressed in the original image. In the inference process, this study used the YOLO-v5 algorithm, known for its fast processing speed and widely used for real-time object detection. This method could effectively detect large and even small objects that were difficult to detect with the existing object detection algorithms. When the large-scale site monitoring system was tested, it performed well in detecting small objects, such as workers in a large-scale view of construction sites, which were inaccurately detected by the existing algorithms. Our next goal is to incorporate various safety monitoring and risk analysis algorithms into this system, such as collision risk estimation, based on the time-to-collision concept, enabling the optimization of safety routes by accumulating workers' paths and inferring the risky areas based on workers' trajectory patterns. Through such developments, this continuous large-scale site monitoring system can guide a construction plant's safety management system more effectively.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.57-73
• /
• 2021

Maintenance and prevention of failure through anomaly detection of ICT infrastructure is becoming important. System monitoring data is multidimensional time series data. When we deal with multidimensional time series data, we have difficulty in considering both characteristics of multidimensional data and characteristics of time series data. When dealing with multidimensional data, correlation between variables should be considered. Existing methods such as probability and linear base, distance base, etc. are degraded due to limitations called the curse of dimensions. In addition, time series data is preprocessed by applying sliding window technique and time series decomposition for self-correlation analysis. These techniques are the cause of increasing the dimension of data, so it is necessary to supplement them. The anomaly detection field is an old research field, and statistical methods and regression analysis were used in the early days. Currently, there are active studies to apply machine learning and artificial neural network technology to this field. Statistically based methods are difficult to apply when data is non-homogeneous, and do not detect local outliers well. The regression analysis method compares the predictive value and the actual value after learning the regression formula based on the parametric statistics and it detects abnormality. Anomaly detection using regression analysis has the disadvantage that the performance is lowered when the model is not solid and the noise or outliers of the data are included. There is a restriction that learning data with noise or outliers should be used. The autoencoder using artificial neural networks is learned to output as similar as possible to input data. It has many advantages compared to existing probability and linear model, cluster analysis, and map learning. It can be applied to data that does not satisfy probability distribution or linear assumption. In addition, it is possible to learn non-mapping without label data for teaching. However, there is a limitation of local outlier identification of multidimensional data in anomaly detection, and there is a problem that the dimension of data is greatly increased due to the characteristics of time series data. In this study, we propose a CMAE (Conditional Multimodal Autoencoder) that enhances the performance of anomaly detection by considering local outliers and time series characteristics. First, we applied Multimodal Autoencoder (MAE) to improve the limitations of local outlier identification of multidimensional data. Multimodals are commonly used to learn different types of inputs, such as voice and image. The different modal shares the bottleneck effect of Autoencoder and it learns correlation. In addition, CAE (Conditional Autoencoder) was used to learn the characteristics of time series data effectively without increasing the dimension of data. In general, conditional input mainly uses category variables, but in this study, time was used as a condition to learn periodicity. The CMAE model proposed in this paper was verified by comparing with the Unimodal Autoencoder (UAE) and Multi-modal Autoencoder (MAE). The restoration performance of Autoencoder for 41 variables was confirmed in the proposed model and the comparison model. The restoration performance is different by variables, and the restoration is normally well operated because the loss value is small for Memory, Disk, and Network modals in all three Autoencoder models. The process modal did not show a significant difference in all three models, and the CPU modal showed excellent performance in CMAE. ROC curve was prepared for the evaluation of anomaly detection performance in the proposed model and the comparison model, and AUC, accuracy, precision, recall, and F1-score were compared. In all indicators, the performance was shown in the order of CMAE, MAE, and AE. Especially, the reproduction rate was 0.9828 for CMAE, which can be confirmed to detect almost most of the abnormalities. The accuracy of the model was also improved and 87.12%, and the F1-score was 0.8883, which is considered to be suitable for anomaly detection. In practical aspect, the proposed model has an additional advantage in addition to performance improvement. The use of techniques such as time series decomposition and sliding windows has the disadvantage of managing unnecessary procedures; and their dimensional increase can cause a decrease in the computational speed in inference.The proposed model has characteristics that are easy to apply to practical tasks such as inference speed and model management.

• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.1-25
• /
• 2020

Sentiment Analysis (SA) is a Natural Language Processing (NLP) task that analyzes the sentiments consumers or the public feel about an arbitrary object from written texts. Furthermore, Aspect-Based Sentiment Analysis (ABSA) is a fine-grained analysis of the sentiments towards each aspect of an object. Since having a more practical value in terms of business, ABSA is drawing attention from both academic and industrial organizations. When there is a review that says "The restaurant is expensive but the food is really fantastic", for example, the general SA evaluates the overall sentiment towards the 'restaurant' as 'positive', while ABSA identifies the restaurant's aspect 'price' as 'negative' and 'food' aspect as 'positive'. Thus, ABSA enables a more specific and effective marketing strategy. In order to perform ABSA, it is necessary to identify what are the aspect terms or aspect categories included in the text, and judge the sentiments towards them. Accordingly, there exist four main areas in ABSA; aspect term extraction, aspect category detection, Aspect Term Sentiment Classification (ATSC), and Aspect Category Sentiment Classification (ACSC). It is usually conducted by extracting aspect terms and then performing ATSC to analyze sentiments for the given aspect terms, or by extracting aspect categories and then performing ACSC to analyze sentiments for the given aspect category. Here, an aspect category is expressed in one or more aspect terms, or indirectly inferred by other words. In the preceding example sentence, 'price' and 'food' are both aspect categories, and the aspect category 'food' is expressed by the aspect term 'food' included in the review. If the review sentence includes 'pasta', 'steak', or 'grilled chicken special', these can all be aspect terms for the aspect category 'food'. As such, an aspect category referred to by one or more specific aspect terms is called an explicit aspect. On the other hand, the aspect category like 'price', which does not have any specific aspect terms but can be indirectly guessed with an emotional word 'expensive,' is called an implicit aspect. So far, the 'aspect category' has been used to avoid confusion about 'aspect term'. From now on, we will consider 'aspect category' and 'aspect' as the same concept and use the word 'aspect' more for convenience. And one thing to note is that ATSC analyzes the sentiment towards given aspect terms, so it deals only with explicit aspects, and ACSC treats not only explicit aspects but also implicit aspects. This study seeks to find answers to the following issues ignored in the previous studies when applying the BERT pre-trained language model to ACSC and derives superior ACSC models. First, is it more effective to reflect the output vector of tokens for aspect categories than to use only the final output vector of [CLS] token as a classification vector? Second, is there any performance difference between QA (Question Answering) and NLI (Natural Language Inference) types in the sentence-pair configuration of input data? Third, is there any performance difference according to the order of sentence including aspect category in the QA or NLI type sentence-pair configuration of input data? To achieve these research objectives, we implemented 12 ACSC models and conducted experiments on 4 English benchmark datasets. As a result, ACSC models that provide performance beyond the existing studies without expanding the training dataset were derived. In addition, it was found that it is more effective to reflect the output vector of the aspect category token than to use only the output vector for the [CLS] token as a classification vector. It was also found that QA type input generally provides better performance than NLI, and the order of the sentence with the aspect category in QA type is irrelevant with performance. There may be some differences depending on the characteristics of the dataset, but when using NLI type sentence-pair input, placing the sentence containing the aspect category second seems to provide better performance. The new methodology for designing the ACSC model used in this study could be similarly applied to other studies such as ATSC.

• Journal of KIISE
• /
• v.42 no.10
• /
• pp.1195-1206
• /
• 2015

For the management of a knowledge system, systems that automatically infer and manage scalable knowledge are required. Most of these systems use ontologies in order to exchange knowledge between machines and infer new knowledge. Therefore, approaches are needed that infer new knowledge for scalable ontology. In this paper, we propose an approach to perform rule based reasoning for scalable SHIF ontologies in a spark framework which works similarly to MapReduce in distributed memories on a cluster. For performing efficient reasoning in distributed memories, we focus on three areas. First, we define a data structure for splitting scalable ontology triples into small sets according to each reasoning rule and loading these triple sets in distributed memories. Second, a rule execution order and iteration conditions based on dependencies and correlations among the SHIF rules are defined. Finally, we explain the operations that are adapted to execute the rules, and these operations are based on reasoning algorithms. In order to evaluate the suggested methods in this paper, we perform an experiment with WebPie, which is a representative ontology reasoner based on a cluster using the LUBM set, which is formal data used to evaluate ontology inference and search speed. Consequently, the proposed approach shows that the throughput is improved by 28,400% (157k/sec) from WebPie(553/sec) with LUBM.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.44 no.6
• /
• pp.19-29
• /
• 2007

This paper introduces a robot knowledge framework which is represented with multiple classes, levels and layers to implement robot intelligence at real environment for mobile robot. Our root knowledge framework consists of four classes of knowledge (KClass), axioms, rules, a hierarchy of three knowledge levels (KLevel) and three ontology layers (OLayer). Four KClasses including perception, model, activity and context class. One type of rules are used in a way of unidirectional reasoning. And, the other types of rules are used in a way of bi-directional reasoning. The robot knowledge framework enable a robot to integrate robot knowledge from levels of its own sensor data and primitive behaviors to levels of symbolic data and contextual information regardless of class of knowledge. With the integrated knowledge, a robot can have any queries not only through unidirectional reasoning between two adjacent layers but also through bidirectional reasoning among several layers even with uncertain and partial information. To verify our robot knowledge framework, several experiments are successfully performed for object recognition and navigation.

• The KIPS Transactions:PartB
• /
• v.8B no.5
• /
• pp.429-440
• /
• 2001

To achieve a computer tutor framework with high learning effects as well as practicality, the goal of this research has been set to developing an intelligent tutor for problem-solving in mathematics domain. The maine feature of the CyberTutor, a computer tutor developed in this research, is the facilitation of a learning environment interacting in accordance with the learners differing inferential capabilities and needs. The pedagogical information, the driving force of such an interactive learning, comprises of tutoring strategies used commonly in various domains such as phvsics and mathematics, in which the main contents of learning is the comprehension and the application of principles. These tutoring strategies are those of testing learners hypotheses test, providing hints, and generating explanations. We illustrate the feasibility and the behavior of our propose framework with a sample problem-solving learning in geometry. The proposed tutorial framework is an advancement from previous works in several aspects. Firstly, it is more practical since it supports handing of a wide range of problem types, including not only proof types but also finding-unkown tpes. Secondly, it is aimed at facilitating a personal tutor environment by adapting to learners of varying capabilities. Finally, learning effects are maximized by its tutorial dialogues which are derived from real-time problem-solving inference instead of from built-in procedures.

• Journal of Microbiology and Biotechnology
• /
• v.14 no.2
• /
• pp.276-283
• /
• 2004

Screening was performed to isolate cellulose-producing microorganisms from the Korean traditional fermented persimmon vinegar. The resulting strain, KJ $145^{T}$, was then taxonomically investigated by phenotypic characterization, particularly chemotaxonomic, and by phylogenetic inference based on a 16S rDNA sequence analysis including other related taxa. Strain KJ $145^{T}$ was found to grow rapidly and form pale white colonies with smooth to rough surfaces on a GYC agar. Strain KJ $145^T$ also produced acetate from ethanol, and was tolerable to 10% ethanol in SM medium. In a static culture, a thick cellulose pellicle was produced, and in GYC broth, the strain grew at temperatures ranging from 28 to $40^\circ{C}$ with an optimum pH of 4.0. The genomic DNA G+C content of strain KJ $145^T$ was 61.9 mol%, and the predominant ubiquinone was Q 10 as the major quinone and Q9 as the minor quinone. The major cellular fatty acids were $C_{16:0}$ and the sum in feature 7 ($C_{18:1}$ w9c, w12t and/or w7c). A 16S rRNA-targeted oligonucleotide probe specific for strain KJ $145^T$was constructed, and the phylogenetic position of the new species was derived from a 16S rDNA-based tree. When comparing the 16S rDNA nucleotide sequences, strain KJ $145^T$ was found to be most closely related to G. hansenii LMG $1527^T$ (99.2%), although KJ $145^T$ was still distinct from G. hansenii LMG $l527^T$ and G. xylinus LMG $1515^T$ in certain phenotypic characteristics. Therefore, on the basis of 16S rDNA sequences and taxonomic characteristics, it is proposed that strain KJ $145^T$ should be placed in the genus Gluconacetobacter as a new species, Gluconacetobacter persimmonis sp. nov., under the type-strain KJ $145^T$ (=KCTC =$10175BP^T$=KCCM=$10354^T$).

• The Korean Journal of Applied Statistics
• /
• v.29 no.1
• /
• pp.85-98
• /
• 2016

This study makes an empirical comparison of various realized volatilities (RVs) in terms of overnight returns. In financial asset markets, during overnight or holidays, no or few trading data are available causing a difficulty in computing RVs for a whole span of a day. A review will be made on several RVs reflecting overnight return variations. The comparison is made for forecast accuracies of several RVs for some financial assets: the US S&P500 index, the US NASDAQ index, the KOSPI (Korean Stock Price Index), and the foreign exchange rate of the Korea won relative to the US dollar. The RV of a day is compared with the square of the next day log-return, which is a proxy for the integrated volatility of the day. The comparison is made by investigating the Mean Absolute Error (MAE) and the Root Mean Square Error (RMSE). Statistical inference of MAE and RMSE is made by applying the model confidence set (MCS) approach and the Diebold-Mariano test. For the three index data, a specific RV emerges as the best one, which addresses overnight return variations by inflating daytime RV.