• Journal of Digital Convergence
• /
• v.12 no.12
• /
• pp.293-301
• /
• 2014

In this paper, we propose a new measure based on the confidence of Bayesian posterior probability so as to reduce unimportant information in the clustering process. Because the performance of clustering is up to selecting the important degree of attributes within the databases, the concept of information entropy is added to posterior probability for attributes discernibility. Hence, The same value of attributes in the confidence of the proposed measure is considerably much less due to the natural logarithm. Therefore posterior probability-based clustering algorithm selects the minimum of attribute reducts and improves the efficiency of clustering. Analysis of the validation of the proposed algorithms compared with others shows their discernibility as well as ability of clustering to handle uncertainty with ACME categorical data.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2002.11a
• /
• pp.445-453
• /
• 2002

Past and current customer behavior is the best predicator of future customer behavior. This paper introduces a procedure on personalized defection detection and prevention for an online game site. The basic idea for our defection detection and prevention is adopted from the observation that potential defectors have a tendency to take a couple of months or weeks to gradually change their behavior (i.e. trim-out their usage volume) before their eventual withdrawal. For this purpose, we suggest a SOM (Self-Organizing Map) based procedure to determine the possible states of customer behavior from past behavior data. Based on this representation of the state of behavior, potential defectors are detected by comparing their monitored trajectories of behavior states with frequent and confident trajectories of past defectors. The key feature of this study includes a defection prevention procedure which recommends the desirable behavior state for the ext period so as to lower the likelihood of defection. The defection prevention procedure can be used to design a marketing campaign on an individual basis because it provides desirable behavior patterns for the next period. The experiments demonstrate that our approach is effective for defection prevention and efficient for defection detection because it predicts potential defectors without deterioration of prediction accuracy compared to that of the MLP (Multi-Layer Perceptron) neural network.

• Proceedings of the KAIS Fall Conference
• /
• 2000.10a
• /
• pp.370-374
• /
• 2000

야구 경기에서는 한 경기에 여러 투수가 등판하게 되는데, 상황에 따라 성격이 다른 투수가 공을 던지게 된다. 이러한 등판 투수의 선정은 감독 고유의 권한이며 감독이 오랜 경험을 통해 승리하기 위해 최적의 투수를 선정하게 된다. 본 논문은 그러한 감독의 경험을 학습하기 위하여 프로야구 경기에서 발생하는 기록 데이터를 데이터마이닝을 이용하여 분석한 후, 앞으로 열릴 경기에 등판할 투수를 미리 예측할 수 있는 방안에 대하여 연구하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2004.11a
• /
• pp.449-452
• /
• 2004

게임을 구성하는 다양하고 방대한 데이터 집합을 K-Means 알고리즘을 통해 물리적 혹은 추상적 객체를 비슷한 객체군으로 그룹화 하는 클러스트링화 하여 데이터마이닝 기법을 통한 데이터 최적화 방안을 모색한다. 이는 네트워크 환경에서의 게임의 안정성과 정형화된 패턴을 벗어나지 못하는 게임에 충분한 게임성을 높일 수 있다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.161-164
• /
• 2022

게임 산업은 기술의 발전과 비대면 서비스 수요 증가로 해마다 발전하고 있다. 본 연구는 사용자들의 수요를 조사하기 위하여 대중성이 가장 높은 온라인 게임 플랫폼에서 이용 시간이 많은 게임 정보를 확인하였다. HTML 파싱(parsing) 라이브러리를 통해 해당 게임들의 리뷰를 크롤링하여 엑셀 파일로 데이터베이스화하였고, 자연어 처리 라이브러리를 활용하여 데이터를 정제하였다. 총 5개 장르에 대하여 분석한 결과 각 장르에 해당하는 대표적인 키워드를 확인할 수 있었다. 취득한 키워드는 범용 시각화 패키지를 활용하여 워드 클라우드 형태로 한눈에 알아볼 수 있도록 시각화하였다.

• Journal of Digital Convergence
• /
• v.15 no.6
• /
• pp.229-236
• /
• 2017

An efficient frequent pattern algorithm is essential for mining association rules as well as many other mining tasks for convergence with its application spread over a very broad spectrum. Models for mining pattern have been proposed using a FP-tree for storing compressed information about frequent patterns. In this paper, we propose a centroid frequent pattern growth algorithm which we called "CAWFP-Growth" that enhances he FP-Growth algorithm by making the center of weights and frequencies for the itemsets. Because the conventional constraint of maximum weighted support is not necessary to maintain the downward closure property, it is more likely to reduce the search time and the information loss of the frequent patterns. The experimental results show that the proposed algorithm achieves better performance than other algorithms without scarifying the accuracy and increasing the processing time via the centroid of the items. The MapReduce framework model is provided to handle large amounts of data via a pseudo-distributed computing environment. In addition, the modeling of the proposed algorithm is required in the fully distributed mode.

• Journal of Digital Convergence
• /
• v.13 no.6
• /
• pp.157-164
• /
• 2015

An efficient retrieval of useful information is a prerequisite of an optimal decision making system. Hence, A research of data mining techniques finding useful patterns from the various forms of data has been progressed with the increase of the application of Big Data for convergence and integration with other industries. Each technique is more likely to have its drawback so that the generalization of retrieving useful information is weak. Another integrated technique is essential for retrieving useful information. In this paper, a uncertainty measure of information is calculated such that algebraic probability is measured by Bayesian theory and then information entropy of the probability is measured. The proposed measure generates the effective reduct set (i.e., reduced set of necessary attributes) and formulating the core of the attribute set. Hence, the optimal decision rules are induced. Through simulation deciding contact lenses, the proposed approach is compared with the equivalence and value-reduct theories. As the result, the proposed is more general than the previous theories in useful decision-making.

• Journal of Digital Convergence
• /
• v.14 no.3
• /
• pp.209-217
• /
• 2016

Recently, the outdoor activities were increased in accordance with economic growth and improved quality of life. In addition, weather and outdoor activities are closely related. Currently, Outdoor Activities decisions are determined by the Korea Meteorological Administrator's forecasts and subjective experience. Therefore, we need the analysis method that can provide a basis for the decision on outdoor activities based on meteorological information. In this paper, we propose an algorithm that can analyze meteorological information to support decision-making outdoor activities. And the algorithm is based on the data mining. In addition, we have constructed a baseball game schedule with automatic weather system's observation data in the training data. We verified the improved performance of the proposed algorithm.

• Journal of Korean Library and Information Science Society
• /
• v.52 no.3
• /
• pp.313-334
• /
• 2021

This study aimed to analyze AI research trends related to elementary, middle, and high school education. To this end, the related literature was collected from the SCOPUS database and the publication period of the collected literature was from 1974 to March 2021, with 154 journal papers and 571 conference papers. Research trends were analyzed based on the co-occurrences analysis technique of 4,521 words of author keyword and index keyword included in these papers. As a result of the analysis, big data, data mining, data science and deep learning were found as the latest research trends with machine learning and there was a difference between elementary, middle and high school education. It can be seen that elementary school had a lot of robot-related research, middle school had a lot of game and data-related research, and high school had various and in-depth research. In discussion, we mapped the top 50 words common to elementary, middle, and high schools with the 'Artificial Intelligence Basics' curriculum of Korean Government and '5 Big Ideas' of the United States Government so that AI research can be viewed at a glance.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.22 no.3
• /
• pp.585-600
• /
• 2012

Security issues in online games are increasing as the online game industry grows. Real money trading (RMT) by online game users has become a security issue in several countries including Korea because RMT is related to criminal activities such as money laundering or tax evasion. RMT-related activities are done by professional work forces, namely gold-farmers, and many of them employ the automated program, bot, to gain cyber asset in a quick and efficient way. Online game companies try to prevent the activities of gold-farmers using game bots detection algorithm and block their accounts or IP addresses. However, game bot detection algorithm can detect a part of gold-farmer's network and IP address blocking also can be detoured easily by using the virtual private server or IP spoofing. In this paper, we propose a method to detect gold-farmer groups by analyzing their connection patterns to the online game servers, particularly information on their routing and source locations. We verified that the proposed method can reveal gold-farmers' group effectively by analyzing real data from the famous MMORPG.