• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.01a
• /
• pp.57-58
• /
• 2019

본 논문에서는 앞서 진행한 연구들과 딥러닝을 이용한 고소작업자 행동 모니터링 논문에 이어 작업자 위험 행동분류 시스템을 개선할 수 있는 연구 결과를 비교, 설명한다. 이번 연구에서는 작업자의 행동에 따른 고도계 센서의 데이터를 추가로 수집하여 작업자의 더 다양한 행동을 분류하고 위험 행동 패턴 분석을 위한 방향을 제시한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2010.11a
• /
• pp.542-545
• /
• 2010

모바일은 사용자의 여섯 번째 감각의 도구로써 신체의 일부가 되고 있으며 사용자의 감정이나 행동패턴 등의 상황 정보를 분석하여 사용자의 의도나 요구사항을 예측할 수가 있다. 본 논문에서는 안드로이드를 기반으로 사용자의 감정을 포함한 다양한 상황을 인지하고 감정과 상황에 따른 사용자의 행동을 추론할 수 있는 상황인지 스마트 폰을 제안한다. 상황인지 스마트폰은 카메라에서 입력된 사용자의 얼굴 영상을 통해 감정을 인지하고 감정에 따른 행동과 그 때의 상황정보를 조합하여 히스토리 DB에 축적된다. 또한 현재의 감정과 상황 정보에 따라 과거의 히스토리DB를 비교 분석하고 현재의 행동을 추론한다. 이것은 사용자에게 고차원적인 서비스를 제공해줄 수 있다. 상황인지 스마트폰은 감정을 인지하는 감정인지 매니저와 상황 정보를 수집하고 관리하는 컨텍스트 매니저, 행동을 추론하는 추론 매니저로 구성된다.

• The Journal of the Korea Contents Association
• /
• v.6 no.1
• /
• pp.64-71
• /
• 2006

Electronic mail (E-mail) is an integral part of communication for the recent Internet users. However, e-mail has also come to serve as a means to support flood of unwanted spam mails and junk mails having bad purposes. This paper was conducted in order to develop an intelligent e-mail system using user behavior pattern that can prevent these unnecessary information and enable the user to enjoy communication via e-mail in a cleaner environment. The concentrated analysis of the user behavior in terms of using e-mail functions has resulted in better classification between unnecessary and necessary information, thereby facilitating faster disposal of spam mails.

• Journal of Digital Contents Society
• /
• v.15 no.5
• /
• pp.579-586
• /
• 2014

This paper analyzes the recognition of driver's behavior in lane change using context-free grammar. In contrast to conventional pattern recognition techniques, context-free grammars are capable of describing features effectively that are not easily represented by finite symbols. Instead of coordinate data processing that should handle features in multiple concurrent events respectively, effective syntactic analysis was applied for patterning of symbolic sequence. The findings proposed the effective and intuitive method for drivers and researchers in driving safety field. Probabilistic parsing for the improving this research will be the future work to achieve a robust recognition.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2015.10a
• /
• pp.501-503
• /
• 2015

최근, 웨어러블 디바이스의 발달로 센서를 활용한 애플리케이션의 분야가 점점 확대되고 있다. 웨어러블 디바이스의 가속도, 자이로 센서를 이용한 걷기, 달리기 등의 행위 판단이 가능해졌다. 이를 바탕으로 본 논문에서는 스마트폰과 웨어러블 디바이스, 두 가지를 활용해 신체 및 손의 특정 행동을 수직, 수평 성분 추출과 데이터마이닝 툴을 이용해 분석한다. 이를 통해 특정 행동의 패턴 판단을 방법을 제시한다.

• Journal of the Institute of Convergence Signal Processing
• /
• v.23 no.2
• /
• pp.91-96
• /
• 2022

At the current camera-based technology level, sensor-based basic life pattern recognition technology has to suffer inconvenience to obtain accurate data, and commercial band products are difficult to collect accurate data, and cannot take into account the motive, cause, and psychological effect of behavior. the current situation. In this paper, radar technology for life pattern recognition is a technology that measures the distance, speed, and angle with an object by transmitting a waveform designed to detect nearby people or objects in daily life and processing the reflected received signal. It was designed to supplement issues such as privacy protection in the existing image-based service by applying it. For the implementation of the proposed system, based on TI IWR1642 chip, RF chipset control for 60GHz band millimeter wave FMCW transmission/reception, module development for distance/speed/angle detection, and technology including signal processing software were implemented. It is expected that analysis of individual life patterns will be possible by calculating self-management and behavior sequences by extracting personalized life patterns through quantitative analysis of life patterns as meta-analysis of living information in security and safe guards application.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2003.05a
• /
• pp.586-590
• /
• 2003

고객 분류는 고객관계관리(CRM)의 한 부분으로서 기업에게 이익을 주는 고객의 속성과 구매패턴을 분석함으로써 목표 고객을 결정하는 것을 의미한다. 현재까지 고객 분류에 관한 연구는 특정한 시점에서 고객의 속성과 구매 패턴을 분석함으로써 이루어졌다. 그러나 인터넷 쇼핑몰과 같은 동적인 환경에 있어서 기존의 정적인 분석방법은 시간에 따라 지속적으로 변하는 고객의 행동 변화를 찾아내고, 예측하는데 적합하지 않다. 본 논문에서는 Decision Tree, ANOVA 분석, ARIMA 모형을 사용하여, 특정한 시점에서의 고객 분류뿐만 아니라 미래 시점에서의 고객 분류를 예측하고 패턴을 분석하는 동적인 고객 분류 방법을 제안한다. 동적인 고객 분류를 통해 인터넷 쇼핑몰 기업은 효율적인 마케팅 전략을 작성하여 기업의 이익을 증진시킬 수 있다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.6
• /
• pp.712-717
• /
• 2011

Recently, diverse schedule management agents are being researched for the efficient schedule management of smart devices users, but they remain at a confirmatory level. In order to efficiently manage user's schedules, execution of planned schedules should be monitored to help users properly execute their schedules, or feedback must be given so that when setting up new schedules, users can plan their schedule according to their schedule establishment patterns. This research proposes a schedule management agent that infers the user's behaviors by using acquired user context, and provides schedule related feedback depending on the user's behavior patterns, when users are executing their schedules or planning new schedules. For this, collected user context information is preprocessed and user's behavior is inferred by Bayesian network. Also, in order to provide feedbacks necessary for confirming the user's schedule execution and new schedule establishment, a context tree pattern matching method for the user's schedule, location and time contexts was applied, then verified with 6 weeks of user simulation in a mobile environment.

• The Journal of the Korea Contents Association
• /
• v.19 no.11
• /
• pp.247-261
• /
• 2019

Demand/withdraw communication is an important dysfunctional pattern of serial arguing. This study aims at addressing factors that affect the ways in which self-demand/partner-withdraw pattern increases the likelihood of persistence of serial arguing. We posit that sensitivity to rejection is positively related to the degree to which individuals perceive a partners' behavior as generally disconfirming, which is positively related to enactment of self-demand/partner-withdraw during an argumentative episode. This sequence is positively related to perception of arguments as unresolved. In addition, among those who reported their argument was resolved, this sequence is positively related to the likelihood that the argument is resolved without mutual agreement. Serial mediation analysis confirmed that the likelihood of resolution without mutual agreement were positively associated with rejection sensitivity partially because high RS individuals are likely to perceive their partner to be generally disconfirming and to enact self-demand/partner-withdraw communication during the episode. However, this pattern did not apply to perception of the argument as resolved.

• The KIPS Transactions:PartD
• /
• v.16D no.3
• /
• pp.423-432
• /
• 2009

Advances in information and communication technologies are changing our society greatly. In knowledge-based society, information can be obtained easily via communication tools such as web and e-mail. However, obtaining right and up-to-date information is difficult in spite of overflowing information. The concept of adaptive web site has been initiated recently. The purpose of the site is to provide information only users want out of tons of data gathered. In this paper, an algorithm is developed for adaptive web site construction. The proposed algorithm is based on association rules that are major principle in adaptive web site construction. The algorithm is constructed by analysing log data in web server and extracting meaning documents through finding behavior patterns of users. The proposed algorithm has the following characteristics. First, it is superior to existing algorithms using association rules in time complexity. Its superiority is proved theoretically. Second, the proposed algorithm is effective in space complexity. This is due to that it does not need any intermediate products except a linked list that is essential for finding frequent item sets.