• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.394-395
• /
• 2023

"이성적이며 이상적인 합리적 인간"을 가정하는 기존 경제학의 이론이 항상 실제 상황과 일치하지는 않는 것으로 알려져 있다. 이의 대안으로 나온 행동경제학은, 인간의 경제적 의사결정에 심리, 인지, 감정, 사회문화적 배경 등이 영향을 미친다고 본다. 본 연구에서는 행동경제학에 의거하여, 개인의 감정과 경험이 경제적 의사결정에 영향을 미치는지 여부를 빅데이터 모델을 활용하여 분석하였다. SNS 여론으로는 Reddit, 주가지수로는 S&P 500 을 선정하였다. 수집한 텍스트 데이터를 전처리와 감정분석을 통해 독립변수 값으로 사용했고, 주가지수 등락의 방향성을 종속변수로 사용하여 로지스틱 모형을 구성했다. 모델을 활용하여 분석한 결과 Public sentiment 와 Market sentiment 간 양의 상관관계를 확인할 수 있었다. 또한, lag 를 설정하는 모델이 정확도가 더욱 높음을 확인해, 기존 경제학의 EMH 와 대립되는 바를 확인할 수 있었다. 하지만 최적의 lag 산정을 위해, 더 광범위한 데이터를 바탕으로 한 후속연구가 필요하다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.498-500
• /
• 2022

CCTV protects people and assets safely by identifying dangerous situations and responding promptly. However, it is difficult to continuously monitor the increasing number of CCTV images. For this reason, there is a need for a device that continuously monitors CCTV images and notifies when abnormal behavior occurs. Recently, many studies using artificial intelligence models for image data analysis have been conducted. This study simultaneously learns spatial and temporal characteristic information between image data to classify various abnormal behaviors that can be observed in CCTV images. As an artificial intelligence model used for learning, we propose a multi-classification deep learning model that combines an end-to-end 3D convolutional neural network(CNN) and ResNet.

• The Journal of Bigdata
• /
• v.2 no.2
• /
• pp.105-114
• /
• 2017

Today, many services are supported on the web sites. Analysis of usage patterns of web site visitors is very important to optimize the use and efficiency of the web sites. In this study, analysis of usage patterns and comparative analysis of user groups were conducted by analyzing web access log provided by BPI Challenge 2016. This data provides access logs to the web site in the IT system of a Dutch Employee Insurance Agency (UWV). The customer information, and the click data describing the customers' behavior when using the agency's web site. In this study, we use process mining techniques to analyze the usage patterns of customers and the characteristics of customer groups, and ultimately improve the service quality of customers using web services.

• KIPS Transactions on Computer and Communication Systems
• /
• v.10 no.10
• /
• pp.277-284
• /
• 2021

Recently, the smart home environment is expected to be a platform that collects, integrates, and utilizes various data through convergence with wireless information and communication technology. In fact, the number of smart devices with various sensors is increasing inside smart homes. The amount of data that needs to be processed by the increased number of smart devices is also increasing, and big data processing systems are actively being introduced to handle it effectively. However, traditional big data processing systems have all requests directed to cluster drivers before they are allocated to distributed nodes, leading to reduced cluster-wide performance sharing as cluster drivers managing segmentation tasks become bottlenecks. In particular, there is a greater delay rate on smart home devices that constantly request small data processing. Thus, in this paper, we design a Apriori-based big data system for effective data processing in smart home environments where frequent requests occur at the same time. According to the performance evaluation results of the proposed system, the data processing time was reduced by up to 38.6% from at least 19.2% compared to the existing system. The reason for this result is related to the type of data being measured. Because the amount of data collected in a smart home environment is large, the use of cache servers plays a major role in data processing, and association analysis with Apriori algorithms stores highly relevant sensor data in the cache.

• Journal of Service Research and Studies
• /
• v.10 no.2
• /
• pp.1-15
• /
• 2020

As big data become more important socially and economically in recent years, many problems have been derived from the indiscriminate application of big data. Big data are valuable because it can figure out the meaning of informative information hidden within the data. In particular, to predict customer behavior patterns and experiences, structured data that were extracted from Customer Relationship Management (CRM) or unstructured data that were extracted from Social Network Service(SNS) can be defined as a population to interpret the data, during which many errors can occur. However, those errors are usually overlooked. In addition to data analysis techniques, some data, which should be considered in the analysis, are not included in the population and thus do not show any meaningful patterns. Therefore, this study presents the measurement and interpretation of the data generated when the cause of error in the population setting is strong relationship and interaction between people or a person and an object. In other words, it will be shown that if the relationship and interaction are strong, it is important to include data collected from the perspective of user experience and ethnography in the population by comparing various cases of big data application, through which the meaning will be derived and the best direction will be suggested.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.1498-1501
• /
• 2013

최근 SNS나 스마트폰을 이용한 다양하고 많은 데이터들이 우리 주위에 산재하고 있다. 이러한 데이터는 사용자의 심리나 상황을 담은 것으로 이에 대한 분석을 통해 사용자의 행동과 심리를 유추할 수 있다. 그러나 이러한 데이터는 빅데이터의 형태를 가지고 있기 때문에 이를 효과적으로 분석하기 위한 방법론이 필요하다. 본 논문에서는 이러한 문제에 초점을 맞추고 이를 효과적으로 분석하기 위하여, 먼저 시스템인 R을 소개하고, R에 실제 데이터를 로딩 하여, 이를 분석하는 분석 예를 보인다.

• The Journal of Bigdata
• /
• v.1 no.1
• /
• pp.67-75
• /
• 2016

The study analyzed big data extracted from Google and social media to identify factors related to searches on cyber bullying in Korea and America. Korea's cyber bullying analysis was conducted social big data collected from online news sites, blogs, $caf{\acute{e}}s$, social network services and message for between January 1, 2011 and March 31, 2013. Google search trends for the search words of stress, exercise, drinking, and cyber bullying were obtained for January 1, 2004 and December 22, 2013. The main results of this study were as follows: first, the significant factors stress were cyber bullying that Korea more than America. Secondly, a positive relationship was found between stress and drinking, exercise and cyber bullying both Korea and America. Thirdly, significant differences were found all path both Korea and America. The study shows that both adults and teenagers are influenced in Korea. We need to develop online application that if cyber bullying behavior was predicted can intervene in real time because these actual cyber bullying-related exposure to psychological and behavioral characteristic.

• Information Systems Review
• /
• v.21 no.2
• /
• pp.1-27
• /
• 2019

It is possible to provide Smart Tourism Service through the development of information technology. It is necessary for the tourism industry to understand and utilize Big Data that has tourists' consumption patterns and service usage patterns in order to continuously create a new business model by converging with other industries. This study suggests to activate Jeju Smart Tourism by analyzing Big Data based on credit card usage records and location of tourists in Jeju. The results of the study show that First, the percentage of Chinese tourists visiting Jeju has decreased because of the effect of THAAD. Second, Consumption pattern of Chinese tourists is mostly occurring in the northern areas where airports and duty-free shops are located, while one in other regions is very low. The regional economy of Jeju City and Seogwipo City shows a overall stagnation, without changes in policy, existing consumption trends and growth rates will continue in line with regional characteristics. Third, we need a policy that young people flow into by building Jeju Multi-complex Mall where they can eat, drink, and go shopping at once because the number of young tourists and the price they spend are increasing. Furthermore, it is necessary to provide services for life-support related to weather, shopping, traffic, and facilities etc. through analyzing Wi-Fi usage location. Based on the results, we suggests the marketing strategies and public policies for understanding Jeju tourists' patterns and stimulating Jeju tourism industry.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.23 no.5
• /
• pp.969-976
• /
• 2013

The information protection systems of company are intended to detect all weak points, intrusion, document drain. All actions of people in company are recorded and can check persistently. On the other hand, what analyze security log generated by these systems becomes more difficult. Most staff who manages the security systems, and analyze log is more incomprehensible than a user or a person of drain for an information distribution process of the work-site operations and the management procedure of the critical information. Such a reality say the serious nature of the internal information leakage that can be brought up more. While the research on the big data proceeds actively recently, the successful cases are being announced in the various areas. This research is going to present the improved big data processing technology and case of the security field.

• Journal of Digital Convergence
• /
• v.15 no.1
• /
• pp.381-391
• /
• 2017

Biometric technology is a technology for authenticating a user using the physical or behavioral features of the inherent characteristics of the individual. With the necessity and efficiency of the technology in the fields of finance, security, access control, medical welfare, inspection, and entertainment, the service range has been expanding. Biometrics using biometric information such as fingerprints and faces have been exposed to counterfeit and disguised threats and become a social problem. Recent studies using a bio-signal from the inside of the body other than the bio-information of the external body are being developed. This paper analyzes the recent research and technology of biometric systems using bio-signals, ECG, heart sounds, EEG, and EMG to present the skills needed for the development direction. In the future, utilizing the deep learning to build and analyze database to manage bio-signal based big data for the complex condition of individuals, biometrics technologies suitable for real time environment are expected to be researched.