• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.249-263
• /
• 2023

Collaborative Filtering, a representative recommendation system methodology, consists of two approaches: neighbor methods and latent factor models. Among these, the latent factor model using matrix factorization decomposes the user-item interaction matrix into two lower-dimensional rectangular matrices, predicting the item's rating through the product of these matrices. Due to the factor vectors inferred from rating patterns capturing user and item characteristics, this method is superior in scalability, accuracy, and flexibility compared to neighbor-based methods. However, it has a fundamental drawback: the need to reflect the diversity of preferences of different individuals for items with no ratings. This limitation leads to repetitive and inaccurate recommendations. The Adaptive Deep Latent Factor Model (ADLFM) was developed to address this issue. This model adaptively learns the preferences for each item by using the item description, which provides a detailed summary and explanation of the item. ADLFM takes in item description as input, calculates latent vectors of the user and item, and presents a method that can reflect personal diversity using an attention score. However, due to the requirement of a dataset that includes item descriptions, the domain that can apply ADLFM is limited, resulting in generalization limitations. This study proposes a Generalized Adaptive Deep Latent Factor Recommendation Model, G-ADLFRM, to improve the limitations of ADLFM. Firstly, we use item ID, commonly used in recommendation systems, as input instead of the item description. Additionally, we apply improved deep learning model structures such as Self-Attention, Multi-head Attention, and Multi-Conv1D. We conducted experiments on various datasets with input and model structure changes. The results showed that when only the input was changed, MAE increased slightly compared to ADLFM due to accompanying information loss, resulting in decreased recommendation performance. However, the average learning speed per epoch significantly improved as the amount of information to be processed decreased. When both the input and the model structure were changed, the best-performing Multi-Conv1d structure showed similar performance to ADLFM, sufficiently counteracting the information loss caused by the input change. We conclude that G-ADLFRM is a new, lightweight, and generalizable model that maintains the performance of the existing ADLFM while enabling fast learning and inference.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.46 no.3
• /
• pp.170-185
• /
• 2023

The advent of big data has brought about the need for analytics. Natural language processing (NLP), a field of big data, has received a lot of attention. Topic modeling among NLP is widely applied to identify key topics in various academic journals. The Korean Society of Industrial and Systems Engineering (KSIE) has published academic journals since 1978. To enhance its status, it is imperative to recognize the diversity of research domains. We have already discovered eight major research topics for papers published by KSIE from 1978 to 1999. As a follow-up study, we aim to identify major topics of research papers published in KSIE from 2000 to 2022. We performed topic modeling on 1,742 research papers during this period by using LDA and BERTopic which has recently attracted attention. BERTopic outperformed LDA by providing a set of coherent topic keywords that can effectively distinguish 36 topics found out this study. In terms of visualization techniques, pyLDAvis presented better two-dimensional scatter plots for the intertopic distance map than BERTopic. However, BERTopic provided much more diverse visualization methods to explore the relevance of 36 topics. BERTopic was also able to classify hot and cold topics by presenting 'topic over time' graphs that can identify topic trends over time.

• Wind and Structures
• /
• v.30 no.4
• /
• pp.367-378
• /
• 2020

Two-dimensional Delayed Detached Eddy Simulation (DDES) was carried out to investigate the uniform flow over a twin-box bridge deck (TBBD) with various gap ratios of L/C=5.1%, 12.8%, 25.6%, 38.5%, 73.3% and 108.2% (L: the gap-width between two girders, C: the chord length of a single girder) at Reynolds number, Re=4×104. The aerodynamic coefficients of the prototype deck with gap ratio of 73.3% obtained from the present simulation were compared with the previous experimental and numerical data for different attack angles to validate the present numerical method. Particular attention is devoted to the fluctuating pressure distribution and forces, shear layer reattachment position, wake velocity and flow pattern in order to understand the effects of gap ratio on dynamic flow interaction with the twin-box bridge deck. The flow structure is sensitive to the gap, thus a change in L/C thus leads to single-side shedding regime at L/C≤25.6%, and co-shedding regime at L/C≥35.8% distinguished by drastic changes in flow structure and vortex shedding. The gap-ratio-dependent Strouhal number gradually increases from 0.12 to 0.27, though the domain frequencies of vortices shedding from two girders are identical. The mean and fluctuating pressure distributions is significantly influenced by the flow pattern, and thus the fluctuating lift force on two girders increases or decreases with increasing of L/C in the single-side shedding and co-shedding regime, respectively. In addition, the flow mechanisms for the variation in aerodynamic performance with respect to gap ratios are discussed in detail.

• Journal of the Korean Ceramic Society
• /
• v.53 no.4
• /
• pp.417-428
• /
• 2016

In this review, an attempt is made to calculate one-dimensional (1-D) electron density profiles from experimentally determined (00l) XRD intensities and possible structural models as well in an effort to understand the collective intracrystalline structures of intercalant molecules of two-dimensional (2-D) nanohybrids with heterostructures. 2-D ceramics, including layered metal oxides and clays, have received much attention due to their potential applicability as catalysts, electrodes, stabilizing agents, and drug delivery systems. 2-D nanohybrids based on such layered ceramics with various heterostructures have been realized through intercalation reactions. In general, the physico-chemical properties of such 2-D nanohybrids are strongly correlated with their heterostructures, but it is not easy to solve the crystal structures due to their low crystallinity and high anisotropic nature. However, the powder X-ray diffraction (XRD) analysis method is thought to be the most powerful means of understanding the interlayer structures of intercalant molecules. If a proper number of well-developed (00l) XRD peaks are available for such 2-D nanohybrids, the 1-D electron density along the crystallographic c-axis can be calculated via a Fourier transform analysis to obtain structural information about the orientations and arrangements of guest species in the interlayer space.

• Proceedings of the KOSOMBE Conference
• /
• v.1997 no.05
• /
• pp.245-248
• /
• 1997

Excessive stress on the bone-stem interface may cause local micromotion that could produce midthigh pain, interface bone resorption and prevent bony ingrowth. It is important for clinician and prosthetic designer to develop an understanding of the load transfer mechanism, its associated stress pattern and its relationships with the particular mechanical characteristics of the femoral stem designs. Finite element method (FEM) is preeminently suited to provide information in this respect. The authors developed 3-dimensional numerical finite element models implanted with the straight stem which is composed of total 1170 elements of 8 nodes and with the curved stem which is composed of total 885 elements of 8 node, and analysed the relative micromotions between the straight stem and the curved stem in immediate postoperative stage of an uncemented total hip replacement in load simulating the single leg stance. The results showed that the rotational displacement was occupied over 90% of total micromotion in both types of stem and was peak at the proximal medial portion of the stem, but markedly less distally. The curved stem was more stable especially in terms of rotational stability. It is recommended that surgeons do not allow the patient weight bearing until bony ingrowth was achieved. In the future more attention should be drawn to increase initial rotational stability of the two types of femoral stem to prevent loosening from excessive micromotion.

• Journal of the Korean Society of Earth Science Education
• /
• v.5 no.1
• /
• pp.105-113
• /
• 2012

The purpose of this study was to examine the effects of storytelling skill on science learning motivation and space perception ability. For this study the 5 grade, 2 class was divided into a research group and a comparative group. The class was pre-tested in order to ensure the same standard. The research group had the science class with storytelling skill, and the comparative group had the class with teacher centered lectures for 10 classes in 10 weeks. The storytelling skill was focused on finding stories, constellation searching, story deciding, story hero deciding, story composition, storytelling completion. To prove the effects of this study, science learning motivation was split up according to attention power, relation, confidence, and sense of satisfaction. Also, space perception ability consisted of two-dimensional rotation, 3 dimension rotations, reflection, three-dimensional searching, number of block, and figure type in pattern. The results of this study are as follows. First, using storytelling skill was effective in science learning motivation. Second, using storytelling skill was effective in space perception ability. Also, after using storytelling skill was good reaction by students. As a result, the elementary science class with storytelling skill had the effects of developing science learning motivation and space perception ability. it means the science class with storytelling skill has potential possibilities and value to develop science learning motivation and space perception ability.

• Journal of the Korean Geographical Society
• /
• v.46 no.3
• /
• pp.331-350
• /
• 2011

Advances in computer technology enabled us to simulate the integrated effects of various geomorphic processes on landscape evolution. This review introduces a theoretical framework for 2-dimensional numerical landscape evolution models (NLEMs) which have recently been used for various research purposes. In particular much attention is paid to the approaches deployed to model major geomorphic processes on a geological time scale in previous research. NLEMs can simulate landscape evolution by numerically solving the partial differential equation which represents the relationship among the geomorphic system components (GSCs). Simple process specifications of the relationships among GSCs on a long-term time scale in terms of quantification and attempts to combine processes represent the initial research on NLEMs. Later researchers have taken these simple NLEMs and elaborated on them. Introducing the theories of NLEMs in this review is expected to help researchers trying to utilize or develop NLEMs.

• Korean Chemical Engineering Research
• /
• v.46 no.4
• /
• pp.647-659
• /
• 2008

Colloids are a heterogeneous system in which particles of a few nanometers to hundreds micrometers in size are finely dispersed in liquid medium, but show homogeneous properties in macroscopic scale. They have attracted much attention not only as model systems of natural atomic and molecular self-assembled structures but also as novel structural materials of practical applications in a wide range of areas. In particular, recent advances in colloidal science have focused on nano-bio materials and devices which are essential for drug discovery and delivery, diagnostics and biomedical applications. In this review, first we introduce nano-bio colloidal systems and surface modification of colloidal particles which creates various functional groups. Then, various methods of fabrication of colloidal particles using holographic lithography, microfluidics and virus templates are discussed in detail. Finally, various applications of colloids in metal inks, three-dimensional photonic crystals and two-dimensional nanopatterns are also reviewed as representative potential applications.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2012.05a
• /
• pp.34-34
• /
• 2012

Heavy storms rainfall has caused many landslides and slope failures especially in the mountainous area of the world. Landslides and slope failures are common geologic hazards and posed serious threats and globally cause billions in monetary losses and thousands of casualies each year so that studies on slope stability and its failure mechanism under rainfall are being increasing attention of these days. Rainfall-induced slope failures are generally caused by the rise in ground water level, and increase in pore water pressures and seepage forces during periods of intense rainfall. The effective stress in the soil will be decreased due to the increased pore pressure, which thus reduces the soil shear strength, eventually resulting in slope failure. During the rainfall, a wetting front goes downward into the slope, resulting in a gradual increase of the water content and a decrease of the negative pore-water pressure. This negative pore-water pressure is referred to as matric suction when referenced to the pore air pressure that contributes to the stability of unsaturated soil slopes. Therefore, the importance is the study of saturated unsaturated soil behaviors in evaluation of slope stability under heavy rainfall condition. In an actual field, a series of failures may occur in a slope due to a rainfall event. So, this study attempts to develop a numerical model to investigate this failure mechanism. A two-dimensional seepage flow model coupled with a one-dimensional surface flow and erosion/deposition model is used for seepage analysis. It is necessary to identify either there is surface runoff produced or not in a soil slope during a rainfall event, while analyzing the seepage and stability of such slopes. Runoff produced by rainfall may result erosion/deposition process on the surface of the slope. The depth of runoff has vital role in the seepage process within the soil domain so that surface flow and erosion/deposition model computes the surface water head of the runoff produced by the rainfall, and erosion/deposition on the surface of the model slope. Pore water pressure and moisture content data obtained by the seepage flow model are then used to analyze the stability of the slope. Spencer method of slope stability analysis is incorporated into dynamic programming to locate the critical slip surface of a general slope.

• The Korean Journal of Applied Statistics
• /
• v.27 no.1
• /
• pp.159-168
• /
• 2014

The news impact curve(NIC) originally proposed by Engle and Ng (1993) is a graphical representation of volatility for financial time series. The NIC is a simple but a powerful tool for identifying variability of a given time series. It is noted that the NIC is suited to symmetric volatility. Recently a lot of attention has been paid to asymmetric volatility models and therefore asymmetric version of the NIC would be useful in the field of financial time series. In this article, we propose to incorporate LASSO in constructing asymmetric NICs based on asymmetric GARCH models. In particular, bilinear GARCH models are considered and illustrated via KOSDAQ data.