• Information Systems Review
• /
• v.23 no.4
• /
• pp.69-86
• /
• 2021

As artificial intelligence technology advances, personalized recommendation systems using big data have attracted huge attention. In the case of beauty products, product preferences are clearly divided depending on customers' skin types and sensitivity along with individual tastes, so it is necessary to provide customized recommendation services based on accumulated customer data. Therefore, by employing deep learning methods, this study proposes a neural network-based recommendation model utilizing both product search history and context information such as gender, skin types and skin worries of customers. The results show that our model with context information outperforms collaborative filtering-based recommender system models using customer search history.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.269-273
• /
• 2008

In ubiquitous environments, clients move between domains freely and its activities in the other domains are growth. Like this environment, the service provider makes access or activity records what they are provided to clients. This record can make a privacy problem to recognize a person or trace some works. So this record must be kept and managed by user instead of the service provider. In this paper, we propose a system that can gather those records from the service provider to home domain server which client's managing by themselves. In addition, if remote domain manager requests that record by the legal process, system can transfer only a range of information which allowed by client to keep personal privacy.

• The Journal of Asian Finance, Economics and Business
• /
• v.6 no.4
• /
• pp.139-145
• /
• 2019

This study examines the dynamic linkages between food price inflation and its volatility in the context of Sri Lanka. The empirical evidence derived from the monthly data for the period from 2003M1 to 2017M12 for Sri Lanka. The relationship between inflation rate and inflation volatility has attracted more attention by theoretical and empirical macroeconomists. Empirical studies on the relationship between food inflation and food inflation variability is scarce in the literature. Food price inflation is defined as log difference of food price series. The volatility of a food price inflation is measured by conditional variance generated by the FIGARCH model. Preliminary analysis showed that food inflation is stationary series. Granger causality test reveals that food inflation seems to exert positive impact on inflation variability. We find no evidence for inflation uncertainty affecting food inflation rates. Hence, the findings of the study supports the Friedman-Ball hypothesis in both cases of consumer food price inflation and wholesale food price inflation. This implies that past information on food inflation can help improve the one-step-ahead prediction of food inflation variability but not vice versa. Our results have some important policy implications for the design of monetary policy, food policy thereby promoting macroeconomic stability.

• Proceedings of the Korean Information Science Society Conference
• /
• 2010.06c
• /
• pp.414-418
• /
• 2010

최근 라이프 로그의 수집과 관리에 관련된 연구가 많이 진행 중에 있다. 또 핸드폰 카메라, 디지털 카메라, 캠코더 등의 발전으로 자신의 일상생활을 비디오로 저장하고, 인터넷을 통해 공유하는 사람도 증가하고 있다. 비디오 데이터는 많은 정보를 포함하고 있는 라이프 로그의 한 예로. 동영상의 촬영 및 수집이 활발해짐에 따라 동영상의 메타정보를 생성하고, 이를 이용해 동영상 검색과 관리에 이용하려는 연구들이 진행 중이다. 본 논문에서는 라이프 로그를 수집하고 수집된 동영상과 라이프 로그를 이용하여 의미정보를 추출하는 시스템을 제안한다. 의미정보란 사용자의 행동을 나타내는 정보로써 컴퓨터 사용, 식사, 집안일, 이동, 외출, 독서, 휴식, 일, 기타로 9가지의 의미정보를 추출한다. 제안하는 방법은 사용자로부터 GPS, 가속도센서, 캠코더를 이용해 실제 데이터를 수집하고, 전처리 과정을 통하여 특징을 추출한다. 이때 추출될 특징은 위치정보와 사용자의 상태정보 그리고 영상처리릍 통한 RGB와 HSL 색공간의 요소와 MPEG-7의 EHD(Edge Histogram Descriptor). CLD(Color Layout Descriptor)이다. 추출된 특징으로부터 사람 행동과 같은 불안정한 상황에서 강점을 보이는 확률모델 네트워크인 베이지안 네트워크를 이용하여 의미정보를 추출한다. 제안하는 방법의 유용성을 보이기 위해 실제 데이터를 수집하고 추론하고 10-Fold Cross-validation을 이용하여 데이터를 검증한다.

• 한국HCI학회:학술대회논문집
• /
• 2008.02a
• /
• pp.573-578
• /
• 2008

Recently, we should control various devices such as TV, audio, DVD player, video player, and set-top box simultaneously to manipulate home theater system. To execute the function the user want in this situation, user should know functions and positions of the buttons in several remote controllers. Normally, people feel difficult due to these realistic problems. Besides, the number of the devices that we can control shall increase, and people will confuse more if the ubiquitous home environment is realized. Therefore, user adaptive interface that provides the summarized functions is required. Moreover there can be a lot of mobile and stationary controller devices in ubiquitous computing environment, so user interface should be adaptive in selecting the functions that user wants and in adjusting the features of UI to fit in specific controller. To implement the user and controller adaptive interface, we modeled the ubiquitous home environment and used modeled context and device information. We have used Bayesian network to get the degree of necessity in each situation. Behavior selection network uses predicted user situation and the degree of necessity, and it selects necessary functions in current situation. Selected functions are used to construct adaptive interface for each controller using presentation template. For experiments, we have implemented ubiquitous home environment and generated controller usage log in this environment. We have confirmed the BN predicted user requirements effectively as evaluating the inferred results of controller necessity based on generated scenario. Finally, comparing the adaptive home UI with the fixed one to 14 subjects, we confirmed that the generated adaptive UI was more useful for general tasks than fixed UI.

• Korean Journal of Construction Engineering and Management
• /
• v.19 no.1
• /
• pp.21-31
• /
• 2018

The rapid increase of the elderly living alone is a critical issue in worldwide as it leads to a rapid increase of a social support costs (e.g., medical expenses) for the elderly. In early stages of dementia, the activities of daily living (ADL) including self-care tasks can be affected by abnormal patterns or behaviors and used as an evidence for the early diagnosis. However, extracting activities using non-intrusive approach is still quite challenging and the existing methods are not fully visualized to understand the behavior pattern or routine. To address these issues, this research suggests a model to extract the activities from coarse-grained data (spatio-temporal data log) and visualize the behavioral context information. Our approach shows the process of extracting and visualizing the subject's spaceactivity map presenting the context of each activity (time, room, duration, sequence, frequency). This research contributes to show a possibility of detecting subject's activities and behavioral patterns using coarse-grained data (limited to spatio-temporal information) with little infringement of personal privacy.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.11
• /
• pp.2577-2596
• /
• 2013

In this paper, we present an Energy-Aware Self-Stabilizing Distributed Clustering protocol based on message-passing model for Ad Hoc networks. The latter does not require any initialization. Starting from an arbitrary configuration, the network converges to a stable state in a finite time. Our contribution is twofold. We firstly give the formal proof that the stabilization is reached after at most n+2 transitions and requires at most $n{\times}log(2n+{\kappa}+3)$ memory space, where n is the number of network nodes and ${\kappa}$ represents the maximum hops number in the clusters. Furthermore, using the OMNeT++ simulator, we perform an evaluation of our approach. Secondly, we propose an adaptation of our solution in the context of Wireless Sensor Networks (WSNs) with energy constraint. We notably show that our protocol can be easily used for constructing clusters according to multiple criteria in the election of cluster-heads, such as nodes' identity, residual energy or degree. We give a comparison under the different election metrics by evaluating their communication cost and energy consumption. Simulation results show that in terms of number of exchanged messages and energy consumption, it is better to use the Highest-ID metric for electing CHs.

• Bioinformatics and Biosystems
• /
• v.1 no.1
• /
• pp.67-72
• /
• 2006

The aim of this paper is to discuss the effect of missing values in detecting differentially expressed genes in a cDNA microarray experiment in the context of a one sample problem. We conducted a cDNA micro array experiment to detect differentially expressed genes for the metastasis of colorectal cancer based on twenty patients who underwent liver resection due to liver metastasis from colorectal cancer. Total RNAs from metastatic liver tumor and adjacent normal liver tissue from a single patient were labeled with cy5 and cy3, respectively, and competitively hybridized to a cDNA microarray with 7775 human genes. We used $M=log_2(R/G)$ for the signal evaluation, where Rand G denoted the fluorescent intensities of Cy5 and Cy3 dyes, respectively. The statistical problem comprises a one sample test of testing E(M)=0 for each gene and involves multiple tests. The twenty cDNA microarray data would comprise a matrix of dimension 7775 by 20, if there were no missing values. However, missing values occur for various reasons. For each gene, the no missing proportion (NMP) was defined to be the proportion of non-missing values out of twenty. In detecting differentially expressed (DE) genes, we used the genes whose NMP is greater than or equal to 0.4 and then sequentially increased NMP by 0.1 for investigating its effect on the detection of DE genes. For each fixed NMP, we imputed the missing values with K-nearest neighbor method (K=10) and applied the nonparametric t-test of Dudoit et al. (2002), SAM by Tusher et al. (2001) and empirical Bayes procedure by $L\ddot{o}nnstedt$ and Speed (2002) to find out the effect of missing values in the final outcome. These three procedures yielded substantially agreeable result in detecting DE genes. Of these three procedures we used SAM for exploring the acceptable NMP level. The result showed that the optimum no missing proportion (NMP) found in this data set turned out to be 80%. It is more desirable to find the optimum level of NMP for each data set by applying the method described in this note, when the plot of (NMP, Number of overlapping genes) shows a turning point.

• Asian Pacific Journal of Cancer Prevention
• /
• v.16 no.6
• /
• pp.2237-2243
• /
• 2015

Background: Adjuvant chemotherapy improves survival in Dukes C colon cancers post-curative resection. However, the evidence for a role with Dukes B lesions remains unproven despite frequent use for disease characterized by poor prognostic features. In view of limited Asia-specific data, this study aimed to determine survival outcomes and identify prognostic factors in a tertiary teaching hospital in Malaysia. Materials and Methods: A total of 116 subjects who underwent curative surgery with and without adjuvant chemotherapy for Duke B and C primary colon adenocarcinomas diagnosed from 2004-2009 were recruited and data were collected retrospectively. Five-year overall survival (OS) and disease free survival (DFS) were analysed using Kaplan-Meier survival analysis and log-rank (Mantel-Cox) test. Prognostic factors were determined using Cox proportional hazards regression with both univariate and multivariate analyses. Results: The survival analysis demonstrated a 5-year OS of 74.0% for all patients, with 74.9% for Dukes C subjects receiving chemotherapy compared to 28.6% in those not receiving chemotherapy (p=0.001). For Dukes B disease, the 5-year survival rate was 82.6% compared to 75.0% for subjects receiving and not receiving chemotherapy, respectively (p=0.17). Independent prognostic factors identified included a CEA level more than 3.5 ng/ml (hazard ratio (HR)=4.78; p=0.008), serosal involvement (HR=3.75; p=0.028) and completion of chemotherapy (HR= 0.20; p=0.007). Conclusions: In a regional context, this study supports current evidence from the West that adjuvant chemotherapy improves survival in Dukes C colon cancers post curative surgery. However, although a clear benefit has yet to be proven for Dukes B disease, our results suggest survival improvement in selected cases.

• Communications for Statistical Applications and Methods
• /
• v.16 no.4
• /
• pp.605-614
• /
• 2009

TGARCH models characterized by asymmetric volatilities have been useful for analyzing various time series in financial econometrics. We are concerned with persistent volatility in the TGARCH context. Park et al. (2009) introduced I-TGARCH process exhibiting a certain persistency in volatility. This article applies I-TGARCH model to various financial time series in Korea and it is obtained that I-TGARCH provides a better fit than competing models.