• Journal of Intelligence and Information Systems
• /
• v.27 no.2
• /
• pp.1-15
• /
• 2021

If it is difficult for us to make decisions, we ask for advice from friends or people around us. When we decide to buy products online, we read anonymous reviews and buy them. With the advent of the Data-driven era, IT technology's development is spilling out many data from individuals to objects. Companies or individuals have accumulated, processed, and analyzed such a large amount of data that they can now make decisions or execute directly using data that used to depend on experts. Nowadays, the recommender system plays a vital role in determining the user's preferences to purchase goods and uses a recommender system to induce clicks on web services (Facebook, Amazon, Netflix, Youtube). For example, Youtube's recommender system, which is used by 1 billion people worldwide every month, includes videos that users like, "like" and videos they watched. Recommended system research is deeply linked to practical business. Therefore, many researchers are interested in building better solutions. Recommender systems use the information obtained from their users to generate recommendations because the development of the provided recommender systems requires information on items that are likely to be preferred by the user. We began to trust patterns and rules derived from data rather than empirical intuition through the recommender systems. The capacity and development of data have led machine learning to develop deep learning. However, such recommender systems are not all solutions. Proceeding with the recommender systems, there should be no scarcity in all data and a sufficient amount. Also, it requires detailed information about the individual. The recommender systems work correctly when these conditions operate. The recommender systems become a complex problem for both consumers and sellers when the interaction log is insufficient. Because the seller's perspective needs to make recommendations at a personal level to the consumer and receive appropriate recommendations with reliable data from the consumer's perspective. In this paper, to improve the accuracy problem for "appropriate recommendation" to consumers, the recommender systems are proposed in combination with context-based deep learning. This research is to combine user-based data to create hybrid Recommender Systems. The hybrid approach developed is not a collaborative type of Recommender Systems, but a collaborative extension that integrates user data with deep learning. Customer review data were used for the data set. Consumers buy products in online shopping malls and then evaluate product reviews. Rating reviews are based on reviews from buyers who have already purchased, giving users confidence before purchasing the product. However, the recommendation system mainly uses scores or ratings rather than reviews to suggest items purchased by many users. In fact, consumer reviews include product opinions and user sentiment that will be spent on evaluation. By incorporating these parts into the study, this paper aims to improve the recommendation system. This study is an algorithm used when individuals have difficulty in selecting an item. Consumer reviews and record patterns made it possible to rely on recommendations appropriately. The algorithm implements a recommendation system through collaborative filtering. This study's predictive accuracy is measured by Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). Netflix is strategically using the referral system in its programs through competitions that reduce RMSE every year, making fair use of predictive accuracy. Research on hybrid recommender systems combining the NLP approach for personalization recommender systems, deep learning base, etc. has been increasing. Among NLP studies, sentiment analysis began to take shape in the mid-2000s as user review data increased. Sentiment analysis is a text classification task based on machine learning. The machine learning-based sentiment analysis has a disadvantage in that it is difficult to identify the review's information expression because it is challenging to consider the text's characteristics. In this study, we propose a deep learning recommender system that utilizes BERT's sentiment analysis by minimizing the disadvantages of machine learning. This study offers a deep learning recommender system that uses BERT's sentiment analysis by reducing the disadvantages of machine learning. The comparison model was performed through a recommender system based on Naive-CF(collaborative filtering), SVD(singular value decomposition)-CF, MF(matrix factorization)-CF, BPR-MF(Bayesian personalized ranking matrix factorization)-CF, LSTM, CNN-LSTM, GRU(Gated Recurrent Units). As a result of the experiment, the recommender system based on BERT was the best.

• Proceedings of the KOR-KST Conference
• /
• 1995.12a
• /
• pp.45-73
• /
• 1995

The current paper presents a system dynamics model which can generate the land use anq transportation system performance simultaneously is proposed. The model system consists of 7 submodels (population, migration of population, household, job growth-employment-land availability, housing development, travel demand, and traffic congestion level), and each of them is designed based on the causality functions and feedback loop structure between a large number of physical, socio-economic, and policy variables. The important advantages of the system dynamics model are as follows. First, the model can address the complex interactions between land use and transportation system performance dynamically. Therefore, it can be an effective tool for evaluating the time-by-time effect of a policy over time horizons. Secondly, the system dynamics model is not relied on the assumption of equilibrium state of urban systems as in conventional models since it determines the state of model components directly through dynamic system simulation. Thirdly, the system dynamics model is very flexible in reflecting new features, such as a policy, a new phenomenon which has not existed in the past, a special event, or a useful concept from other methodology, since it consists of a lots of separated equations. In Chapter I, II, and III, overall approach and structure of the model system are discussed with causal-loop diagrams and major equations. In Chapter V _, the performance of the developed model is applied to the analysis of the impact of highway capacity expansion on land use for the area of Montgomery County, MD. The year-by-year impacts of highway capacity expansion on congestion level and land use are analyzed with some possible scenarios for the highway capacity expansion. This is a first comprehensive attempt to use dynamic system simulation modeling in simultaneous treatment of land use and transportation system interactions. The model structure is not very elaborate mainly due to the problem of the availability of behavioral data, but the model performance results indicate that the proposed approach can be a promising one in dealing comprehensively with complicated urban land use/transportation system.

• Journal of Energy Engineering
• /
• v.23 no.2
• /
• pp.125-148
• /
• 2014

This paper performs a historical analysis on the various factors contributing to the current carbon lock-in of Korean electricity industry by using techo-institutional complex. The possibilities of the industry's carbon lock-out toward more sustainable development are also investigated. It turns out that market, firm, consumer, and government factors are all responsible for the development of the carbon lock-in of Korean power industry; the Korean government consistently favoring large power plants based on the economy of scale; below-cost electricity tariff; inflation policy to suppress increases in power price; rapid demand growth in summer and winter seasons; rigidities of electricity tariff; and expansion of gas-fired and imported coal-fired large power plants. On the other hand, except for nuclear power generation and smart grid, environment laws and new and renewable energy laws are the other remaining factors contributing to the carbon lock-out. Considering three key points that Korea is an export-oriented economy, the generation mix is the most critical factor to decide the amounts of carbon emission in the power industry, and the share of industry and commercial power consumption is over 85%, it is unlikely that Korea will achieve the carbon lock-out of power industry in the near future. Therefore, there are needs for more integrated approaches from market, firm, consumer, and government all together in order to achieve the carbon lock-out in the electricity industry. Firstly, from the market perspective, it is necessary to persue more active new and renewable energy penetration and to guarantee consumer choices by mitigating the incumbent's monopoly power as in the OECD countries. Secondly, from the firm perspective, the promotion of distributed energy system is urgent, which includes new and renewable resources and demand resources. Thirdly, from the consumer perspective, more green choices in the power tariff and customer awareness on the carbon lock-out are needed. Lastly, the government shall urgently improve power planning frameworks to include the various externalities that were not properly reflected in the past such as environmental and social conflict costs.

• Journal of Life Science
• /
• v.26 no.8
• /
• pp.904-910
• /
• 2016

Complex traits are determined by the combined effects of many loci and are affected by gene networks or biological pathways. Systems biology approaches have an important role in the identification of candidate genes related to complex diseases or traits at the system level. The gene network analysis has been performed by diverse types of methods such as gene co-expression, gene regulatory relationships, protein-protein interaction (PPI) and genetic networks. Moreover, the network-based methods were described for predicting gene functions such as graph theoretic method, neighborhood counting based methods and weighted function. However, there are a limited number of researches in livestock. The present study systemically analyzed genes associated with 102 types of economic traits based on the Animal Trait Ontology (ATO) and identified their relationships based on the gene co-expression network and PPI network in cattle. Then, we constructed the two types of gene network databases and network visualization system (http://www.nabc.go.kr/cg). We used a gene co-expression network analysis from the bovine expression value of bovine genes to generate gene co-expression network. PPI network was constructed from Human protein reference database based on the orthologous relationship between human and cattle. Finally, candidate genes and their network relationships were identified in each trait. They were typologically centered with large degree and betweenness centrality (BC) value in the gene network. The ontle program was applied to generate the database and to visualize the gene network results. This information would serve as valuable resources for exploiting genomic functions that influence economically and agriculturally important traits in cattle.

• Journal of the Korea Concrete Institute
• /
• v.21 no.3
• /
• pp.327-335
• /
• 2009

In the seismic region, non-ductile structures often form soft story and exhibit brittle collapse. However, structure demolition and new structure construction strategies have serious problems, as construction waste, environmental pollution and popular complain. And these methods can be uneconomical. Therefore, to satisfy seismic performance, so many seismic retrofit methods have been investigated. There are some retrofit methods as infill walls, steel brace, continuous walls, buttress, wing walls, jacketing of column or beam. Among them, the infilled frames exhibit complex behavior as follows: flexible frames experiment large deflection and rotations at the joints, and infilled shear walls fail mainly in shear at relatively small displacements. Therefore, the combined action of the composite system differs significantly from that of the frame or wall alone. Purpose of research is evaluation on the seismic performance of infill walls, and improvement concept of this paper is use of SHCCs (strain-hardening cementitious composites) to absorb damage energy effectively. The experimental investigation consisted of cyclic loading tests on 1/3-scale models of infill walls. The experimental results, as expected, show that the multiple crack pattern, strength, and energy dissipation capacity are superior for SHCC infill wall due to bridging of fibers and stress redistribution in cement matrix.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.19 no.3
• /
• pp.164-173
• /
• 2017

Alternative cropping systems would be one of climate change adaptation options. Suitable areas for a crop could be identified using a climate suitability model. The EcoCrop model has been used to assess climate suitability of crops using monthly climate surfaces, e.g., the digital climate map at high spatial resolution. Still, a high-performance computing approach would be needed for assessment of climate suitability to take into account a complex terrain in Korea, which requires considerably large climate data sets. The objectives of this study were to implement a script for R, which is an open source statistics analysis platform, in order to use the EcoCrop model under a parallel computing environment and to assess climate suitability of maize using digital climate maps at high spatial resolution, e.g., 1 km. The total running time reduced as the number of CPU (Central Processing Unit) core increased although the speedup with increasing number of CPU cores was not linear. For example, the wall clock time for assessing climate suitability index at 1 km spatial resolution reduced by 90% with 16 CPU cores. However, it took about 1.5 time to compute climate suitability index compared with a theoretical time for the given number of CPU. Implementation of climate suitability assessment system based on the MPI (Message Passing Interface) would allow support for the digital climate map at ultra-high spatial resolution, e.g., 30m, which would help site-specific design of cropping system for climate change adaptation.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.37 no.6
• /
• pp.499-506
• /
• 2019

NDSM (Normalized Digital Surface Model) is key information for the detailed analysis of remote sensing data. Although NDSM can be simply obtained by subtracting a DTM (Digital Terrain Model) from a DSM (Digital Surface Model), in case of UAV (Unmanned Aerial Vehicle) data, it is difficult to get an accurate DTM due to high resolution characteristics of UAV data containing a large number of complex objects on the ground such as vegetation and urban structures. In this study, RGB-based UAV vegetation index, ExG (Excess Green) was used to extract initial seed points having low ExG values for region growing such that a DTM can be generated cost-effectively based on high resolution UAV data. For this process, local window analysis was applied to resolve the problem of erroneous seed point extraction from local low ExG points. Using the DSM values of seed points, region growing was applied to merge neighboring terrain pixels. Slope criteria were adopted for the region growing process and the seed points were determined as terrain points in case the size of segments is larger than 0.25 ㎡. Various slope criteria were tested to derive the optimized value for UAV data-based NDSM generation. Finally, the extracted terrain points were evaluated and interpolation was performed using the terrain points to generate an NDSM. The proposed method was applied to agricultural area in order to extract the above ground heights of crops and check feasibility of agricultural monitoring.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 1999.03a
• /
• pp.175-186
• /
• 1999

Detecting the features of significant patterns from their own historical data is so much crucial to good performance specially in time-series forecasting. Recently, a new data filtering method (or multi-scale decomposition) such as wavelet analysis is considered more useful for handling the time-series that contain strong quasi-cyclical components than other methods. The reason is that wavelet analysis theoretically makes much better local information according to different time intervals from the filtered data. Wavelets can process information effectively at different scales. This implies inherent support for multiresolution analysis, which correlates with time series that exhibit self-similar behavior across different time scales. The specific local properties of wavelets can for example be particularly useful to describe signals with sharp spiky, discontinuous or fractal structure in financial markets based on chaos theory and also allows the removal of noise-dependent high frequencies, while conserving the signal bearing high frequency terms of the signal. To data, the existing studies related to wavelet analysis are increasingly being applied to many different fields. In this study, we focus on several wavelet thresholding criteria or techniques to support multi-signal decomposition methods for financial time series forecasting and apply to forecast Korean Won / U.S. Dollar currency market as a case study. One of the most important problems that has to be solved with the application of the filtering is the correct choice of the filter types and the filter parameters. If the threshold is too small or too large then the wavelet shrinkage estimator will tend to overfit or underfit the data. It is often selected arbitrarily or by adopting a certain theoretical or statistical criteria. Recently, new and versatile techniques have been introduced related to that problem. Our study is to analyze thresholding or filtering methods based on wavelet analysis that use multi-signal decomposition algorithms within the neural network architectures specially in complex financial markets. Secondly, through the comparison with different filtering techniques results we introduce the present different filtering criteria of wavelet analysis to support the neural network learning optimization and analyze the critical issues related to the optimal filter design problems in wavelet analysis. That is, those issues include finding the optimal filter parameter to extract significant input features for the forecasting model. Finally, from existing theory or experimental viewpoint concerning the criteria of wavelets thresholding parameters we propose the design of the optimal wavelet for representing a given signal useful in forecasting models, specially a well known neural network models.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.4
• /
• pp.419-427
• /
• 2020

In the case of large combined weapon systems, such as submarines, the application, and verification of methods of setting the reliability, availability, and maintainability (RAM) target values for conventional weapon systems are limited. Submarines are complex weapon systems with the characteristics of the diversity of operation mode summary and mission profiles (OMS/MP) as well as equipment complexity because they are composed of multiple weapon systems, such as sonar systems and armed systems. Therefore, this study analyzed the development cases of existing weapon systems, i.e., the RAM target value-setting cases, and derived the problems and limitations of the cases to present measures to improve the setting and verification of the ram target values of submarines. In addition, submarines operate around the world and have different operating and maintenance conditions. Therefore, a submarine's ram target values should be set and verified centering on the mission essential equipment and mission critical equipment, instead of all components that constitute weapon systems. This study examined a method to verify the required performance RAM target-value setting, considering the characteristics of submarines as well as the physical performance requirements for the systems and equipment of submarines that must be considered when implementing national defense acquisition projects for submarines.

• Asia pacific journal of information systems
• /
• v.22 no.1
• /
• pp.53-77
• /
• 2012

Financial firms, especially large scaled firms such as KB bank, NH bank, Samsung Card, Hana SK Card, Hyundai Capital, Shinhan Card, etc. should be securely dealing with the personal financial information. Indeed, people have tended to believe that those big financial companies are relatively safer in terms of information security than typical small and medium sized firms in other industries. However, the recent incidents of personal information privacy invasion showed that this may not be true. Financial firms have increased the investment of information protection and security, and they are trying to prevent the information privacy invasion accidents by doing all the necessary efforts. This paper studies how effectively a financial firm will be able to avoid personal financial information privacy invasion that may be deliberately caused by internal staffs. Although there are several literatures relating to information security, to our knowledge, this is the first study to focus on the behavior of internal staffs. The big financial firms are doing variety of information security activities to protect personal information. This study is to confirm what types of such activities actually work well. The primary research model of this paper is based on Theory of Planned Behavior (TPB) that describes the rational choice of human behavior. Also, a variety of activities to protect the personal information of financial firms, especially credit card companies with the most customer information, were modeled by the four-step process Security Action Cycle (SAC) that Straub and Welke (1998) claimed. Through this proposed conceptual research model, we study whether information security activities of each step could suppress personal information abuse. Also, by measuring the morality of internal staffs, we checked whether the act of information privacy invasion caused by internal staff is in fact a serious criminal behavior or just a kind of unethical behavior. In addition, we also checked whether there was the cognition difference of the moral level between internal staffs and the customers. Research subjects were customer call center operators in one of the big credit card company. We have used multiple regression analysis. Our results showed that the punishment of the remedy activities, among the firm's information security activities, had the most obvious effects of preventing the information abuse (or privacy invasion) by internal staff. Somewhat effective tools were the prevention activities that limited the physical accessibility of non-authorities to the system of customers' personal information database. Some examples of the prevention activities are to make the procedure of access rights complex and to enhance security instrument. We also found that 'the unnecessary information searches out of work' as the behavior of information abuse occurred frequently by internal staffs. They perceived these behaviors somewhat minor criminal or just unethical action rather than a serious criminal behavior. Also, there existed the big cognition difference of the moral level between internal staffs and the public (customers). Based on the findings of our research, we should expect that this paper help practically to prevent privacy invasion and to protect personal information properly by raising the effectiveness of information security activities of finance firms. Also, we expect that our suggestions can be utilized to effectively improve personnel management and to cope with internal security threats in the overall information security management system.