Earticle

Many researches have been focused on the analysis of industry disasters in order to reduce them. As a similar endeavor, this paper provides a propensity analysis of injured people from various industries using classification and regression tree(CART), a data mining algorithm. The sample for this work was chosen from 25,157data related to various industries during one year ( 2003.2~2004.1) at Kangwon-Do in Korea. For the purpose of this paper, eight independent variables (injured date, injured time, injured month, type of Injured person, continuous service period, sex, company size, age)are taken from injured person group. According to the analysis result, it is found that five out of the eight factors that are predicted as significant have salient effects. Factors of season, time/hour, day of the week, or month which disasters happened do not show any significant effect. This paper provides common features of injured people. The provided analysis result will be helpful as a starting point for root cause analysis and reduction of industry disasters and also for development of a guideline of safety management.

농구대회는 선수들이 다양한 기본적인 농구 기술을 사용하여 특정한 전술적 조직 형태에 따라 공격과 수비를 바꾸 는 과정이다. 스포츠 분야에서는, 스포츠 훈련, 실제 격투 경기, 학교 스포츠 관리, 국민 체력 테스트에서 많은 양의 데이터 와 정보가 생성될 것이다. 농구 경기 과정에서 코치와 농구 관련 직원은 다양한 수단을 사용하여 라이벌 팀으로부터 데이터 를 수집하게 되는데, 일부는 직관적인 반면, 다른 팀들은 중요한 정보를 직접적으로 표시하지 못할 수도 있다. 데이터 마이닝은 복잡하고 수많은 데이터 리소스에서 현실과 일치하는 유용한 정보를 찾고 그 안에 숨겨진 정보와 지식을 추출하 는 과정입니다. 본 논문에서는 데이터 수집 및 적용과 농구 기술 조치의 사전 처리를 통해 데이터 마이닝 방법을 기반으로 농구 기술 조치 간의 상관 관계를 연구하였다.

Basketball competition is a process in which athletes use various basic basketball techniques to change offense and defense according to certain tactical organizational forms. In the field of sports, a large amount of data and information will be generated in sports training, actual combat competitions, school sports management and national physical fitness tests. In the process of basketball matches, coaches and basketball-related staff will use various means to collect data from rival teams, some of which are intuitive, while others may not be able to directly display their important information. Data mining is the process of finding useful information that is consistent with the reality from the complex and numerous data resources, and extracting the information and knowledge that are hidden in it. In this paper, through the analysis and application of data collection and preprocessing of basketball technical action, based on the data mining method, the correlation between basketball technical action is studied.

최근 들어 저장장치의 발전과 네트워크의 발달로 인하여 대용량의 데이터에 내재되어 있는 정보를 빠른 시간 내에 처리하여 새로운 지식을 창출하려는 요구가 증가하고 있다. 연속적이고 빠르게 증가하는 데이터를 지칭하는 데이터 스트림에서 데이터 마이닝 기법을 이용하여 시간이 흐름에 따라 변하고, 무한적으로 증가하는 데이터 스트림에서의 빈발항목을 찾는 연구가 활발하게 진행되고 있다. 하지만 기존의 연구들은 시간의 흐름에 따른 빈발항목 탐색방법을 적절히 제시하지 못하고 있으며 단지 집계를 이용하여 빈발항목을 탐색하고 있다. 본 논문에서는 데이터 스트림에서 시간적 측면을 고려하여 상대적인 빈발항목을 탐색하기 위한 새로운 알고리즘으로 한정적인 메모리를 고려하여 빈발항목과 부분 빈발항목만을 저장하고 시간의 흐름에 따른 빈발항목의 갱신방법에 관하여 제안하였다. 논문에서 제안하는 알고리즘의 성능은 다양한 실험을 통해서 검증된다. 제안된 방법은 웹 코스웨어로 학습하는 학생들의 행동패턴을 시간대별로 파악하여 빈발항목 및 상대적인 빈발항목을 탐색함으로써 학생들의 학습효과 증진 및 지도 방향을 설정하는데 활용할 수 있다.

Recently, due to technical improvements of storage devices and networks, the amount of data increase rapidly. In addition, it is required to find the knowledge embedded in a data stream as fast as possible. Huge data in a data stream are created continuously and changed fast. Various algorithms for finding frequent itemsets in a data stream are actively proposed. Current researches do not offer appropriate method to find frequent itemsets in which flow of time is reflected but provide only frequent items using total aggregation values. In this paper we proposes a novel algorithm for finding the relative frequent itemsets according to the time in a data stream. We also propose the method to save frequent items and sub-frequent items in order to take limited memory into account and the method to update time variant frequent items. The performance of the proposed method is analyzed through a series of experiments. The proposed method can search both frequent itemsets and relative frequent itemsets only using the action patterns of the students at each time slot. Thus, our method can enhance the effectiveness of learning and make the best plan for individual learning.

Nowadays most colleges are confronting with a serious problem because many students have left their majors at the colleges. In order to make a countermeasure for reducing major separation rate, many universities are trying to find a proper solution. As a similar endeavor, the objective of this paper Is to find a predicting model of students leaving their majors. The sample for this study was chosen from a university in Kangwon-Do during seven years(2000.3.1 2006. 6.30). In this study, the ratio of training sample versus testing sample among partition data was controlled as 50% : 50% for a validation test of data division. Also, this study provides values about accuracy, sensitivity, specificity about three kinds of algorithms including CHAID, CART and C4.5. In addition, ROC chart and gains chart were used for classification of students leaving their majors. The analysis results were very informative since those enable us to know the most important factors such as semester taking a course, grade on cultural subjects, scholarship, grade on majors, and total completion of courses which can affect students leaving their majors.

Recently most universities are suffering from students leaving their majors. In order to make a countermeasure for reducing major separation rate, many universities are trying to find a proper solution. As a similar endeavor, this paper uses decision tree algorithm which is one of the data mining techniques which conduct grouping or prediction into several sub-groups from interested groups. This technique can analyze a feature of type on students leaving their majors. The dataset consists of 5,115 features through data selection from total data of 13,346 collected from a university in Kangwon-Do during seven years(2000.3.1 2006.6.30). The main objective of this study is to evaluate performance of algorithms including CHAID, CART and C4.5 for classification of students leaving their majors with ROC Chart, Lift Chart and Gains Chart. Also, this study provides values about accuracy, sensitivity, specificity using classification table. According to the analysis result, CART showed the best performance for classification of students leaving their majors.

현재 스마트팩토리에 관한 연구는 구축 방안이나 설립 시 고려사항 등에 대해 꾸준히 진행되고 있다. 그러나 스마트 팩토리를 도입한 기업에 대해서는 다양한 연구가 이루어지지 않고 있다. 이 연구에서는 스마트팩토리의 기초단계를 적용한 중소기업을 대상으로 설문조사를 실시하였다. 만족도의 특성을 확인하기 위해 군집분석을 하였고, 만족도에 따라 어떠한 특성을 가지는지 확인하기 위해 의사결정나무와 나이브베이즈 분석을 하였다. 군집분석 결과 만족도가 높은 그룹과 낮은 그룹으로 나뉘는 것을 확인하였으며, 의사결정나무와 나이브베이즈 분석을 실시한 결과 만족도가 높을수록 생산성 개선 정 도가 높은 것을 확인하였다.

Currently, research on smart factories is steadily being carried out in terms of implementation strategies and considerations in construction. Various studies have not been conducted on companies that introduced smart factories. This study conducted a questionnaire survey for SMEs applying the basic stage of smart factory. And the cluster analysis was conducted to examine the characteristics of the company. In addition, we conducted Decision Tree and Naive Bay to examine how the characteristics of a company are derived and compare the results. As a result of the cluster analysis, it was confirmed that the group was divided into the high satisfaction group and the low satisfaction group. The decision tree and the Naive Bay analysis showed that the higher satisfaction group has high productivity.

본 연구는 데이터마이닝 기법을 적용하여 시각장애인의 취업결정요인를 분석·제공함으로써, 장애인의 취업 성공률을 높임과 동시에 정부의 직업재활 개입의 우선순위를 정하여 효율성을 극대화 할 수 있는 방안을 제시하는데 있다. 분석 자료는 2008년 장애인고용패널조사의 제1차년도이며, 전체 패널 데이터 중 시각장애인이면서 전체연령 20세 이상 65세 이하의 590명 중 취업여부에 대한 무응답자102명을 제외한 나머지 488명을 의사결정나무 기법의 하나인 Exhaustive CHAID 알고리즘을 적용하여 분석하였다. 연구결과 국민기초생활수급 여부가 시각장애인의 취업을 예측하게 하는데 있어 결정적인 역할을 하는 것으로 분석되었다. 또한 국민기초생활수급 여부, 직업전교육훈련, 성별의 3가지 요인이 시각장애인의 취업에 개별적으로 작용하는 효과보다는 상호작용 및 보완을 통하였을 때 시너지 효과가 큰 것으로 나타났다. 이러한 결과를 바탕으로 시각장애인의 고용을 증진시키고 직업재활개입의 효과성을 높이기 위해서는 첫째, 시각장애인에게 고용서비스를 제공할 때 국민기초생활 수급자, 직업전교육훈련 경험자, 여성과 같은취업성공률에 긍정적인 요인을 중복하여 가지고 있는 장애인에 대한 발굴 및 우선적인평가 그리고 집중적인 재활개입이 필요할 것으로 판단된다. 둘째, 미혼상태의 시각장애인이 가정 내에서 받지 못하는 사회·심리적인 지지를 받을 수 있는 토대를 만들어 구직욕구에 대한 동기 부여의 기회를 제공해야 된다. 마지막으로 장애인복지관은 시각장애의특성상 이동의 제한으로 직업재활서비스의 접근이 어려운 점을 감안하여 직접 찾아가는서비스를 제공하는 방안을 모색하여야 하는 것으로 제시되었다.

The purpose of this study is to analyze and provide employment decision factor of the visually impaired by applying data mining technique in order to present a plan for increasing the rate of successful employment of the disabled,as well as setting the priority order of the government's intervention in job rehabilitation to maximize the efficiency. As for the analysis of data, the first year data of Panel Survey of Employment for the Disabled in 2008 was used. Among the entire panel data, 488 persons with visual impairment were chosen out of 590 persons between the age of 20and 65 by excluding 102 persons that did not respond on their employment status. Accordingly, one of the decision-making tree techniques Exhaustive CHAID algorithm was applied for the analysis. The analysis result showed that the status on the national basic livelihood security played a determining role in predicting the employment of the visually impaired. In addition, it was found that the three factors of the status on the national basic livelihood security, pre-employment education & training and gender created bigger synergy effect when they inter-complemented one another than when they worked individually in the employment of the visually impaired. For the purpose of enhancing the employment of the visually impaired and increasing the effectiveness of job rehabilitation intervention based on such result, first, it would be necessary when providing employment service to the visually impaired to identify, evaluate first and provide focused rehabilitation intervention for the disabled that have several of the positive factors on the successful employment such as the national basic livelihood security recipient,pre-employment education & training participant and women. Second, it is necessary to build a foundation through which the unmarried visually impaired can received the social & psychological support that they cannot receive at home in order to provide opportunities of giving motivation for the desire for finding a job. Lastly, it was presented that it would be necessary to come up with a plan for providing services to the visually impaired at their locations considering the difficult accessibility of job rehabilitation service due to their mobile restriction according to the characteristics of visual impairment.

It is urgent to prepare countermeasures for traffic congestion problems of Korea's metropolitan area where central functions such as economic, social, cultural, and education are excessively concentrated. Most users of public transportation in metropolitan areas including Seoul use the traffic cards. If various information is extracted from traffic big data produced by the traffic cards, they can provide basic data for transport policies, land usages, or facility plans. Therefore, in this study, we extract valuable information such as the subway passengers' frequent travel patterns from the big traffic data provided by the Seoul Metropolitan Government Big Data Campus. For this, we use a Hadoop (High-Availability Distributed Object- Oriented Platform) to preprocess the big data and store it into a Mongo database in order to analyze it by a sequential pattern data mining technique. Since we analysis the actual big data, that is, the traffic cards' data provided by the Seoul Metropolitan Government Big Data Campus, the analyzed results can be used as an important referenced data when the Seoul government makes a plan about the metropolitan traffic policies.

분산형 데이터마이닝을 위해 의사결정나무 알고리즘은 분산형 협업 환경에 적합하도록 변환되어야 한다. 본 논문에서 제시된 분산형 데이터마이닝 시스템은 각각의 사이트에서 부분적인 데이터를 위한 데이터마이닝 작업을 수행할 수 있는 에이전트와 여러 에이전트들의 협업을 통해 최종적인 의사결정나무 모델을 완성할 수 있도록 에이전트들 간의 통신을 중재하는 미디에이터로 구성되어 있다. 분산형 데이터마이닝의 장점 중에 하나는 여러 사이트에 분산되어 있는 대량의 데이터를 분산 처리하므로 데이터마이닝의 소요시간을 현저하게 줄일 수 있다는 점이다. 그러나 각 사이트들에 존재하고 있는 에이전트들 간의 통신에 부하가 과도하게 걸린다면, 효율적인 시스템으로의 활용도가 낮아질 것 이다. 본 논문은 에이전트들 간에 의사결정나무 모델의 전송량을 최소로 할 수 있는 방법론에 초점을 맞추었다.

A decision tree algorithm should be modified to be suitable in distributed and collaborative environments for distributed data mining. The distributed data mining system proposed in this paper consists of several agents and a mediator. Each agent deals with a local data mining for data in each local site and communicates with one another to build the global decision tree model. The mediator helps several agents to efficiently communicate among them. One of advantages in distributed data mining is to save much time to analyze huge data with several agents. The paper focuses on a transfer technique among agents dealing with each local decision tree model to reduce huge overhead in communication among them.

본 연구는 출소 후 법무보호복지공단에서 숙식제공 서비스를 받는 대상자의 심리상태를 파악함과 더불어, 이를 6개월 이상 유지한 대상자와 6개월 미만에서 중도 포기한 대상자 간 차이가 있는지 확인하고자 하였다. 이에 2011년 9월 5일부터 공단에서 숙식제공 서비스 를 받은 대상자의 심리상담 데이터 270,233건에 대해 Python3.6.9와 Ucinet 프로그램을 활 용해 분석하였다. 해당 데이터는 비정형 및 반정형 데이터로 구성되어 있으므로 텍스트 마이닝 기법을 활용해 분석하였으며, 상담 직원에 의해 기록된 데이터이기 때문에 내담자 와 상담 직원의 용어를 분류하여 분석하였다. 그 결과, 숙식제공 서비스를 받는 대상자들에 게서는 ‘취업’키워드가 가장 높게 나타났으며, CONCOR분석에 따라 출소자의 심리 키워 드는 ‘취업 및 자립 의지’, ‘자리 관리 및 미래지향적 태도’ ‘부정적인 태도 및 대인관계 문제’ , ‘단기적 우려 및 지속 의지의 부재’ 등으로 분류되었다. 이를 토대로 숙식제공 수료 자와 중도포기자를 분석한 결과, 서비스를 6개월 이상 유지한 수료자들은 ‘취업 및 자립 의지’, ‘자기관리 및 미래지향적 태도’에 관한 키워드가 높게 나타났으며, 6개월 미만의 중 도포기자들은 ‘부정적인 태도 및 대인관계 문제’, ‘단편적 걱정에 집중 및 지속 의지의 부재’ 와 관련된 키워드가 많은 것으로 나타나, 두 집단 간의 명확한 차이가 있음이 확인되었다.

This study sought to identify the psychological status of those who received board and lodging services from the Korea Rehabilitation Agency after being released from prison and to see if there was a difference between those who maintained the basic period of provision for more than six months and those who gave up for less than six months. For this purpose, psychological counseling data 270,233 cases of those who received board and lodging services from the agency were analyzed using Python 3.6.9 and ucinet programs from September 5, 2011. Because the data consisted of unstructured and semi-static data, it was analyzed using text mining techniques, and because it was recorded by the staff, the terms of the counselee and the staff were classified and analyzed. As a result, the 'employment' keyword appear at the highest level among those who received the board and lodging service, and the CONCOR analysis showed that the psychological keywords of the ex-offenders were classified as 'employment and self-reliance will', 'self-management and future-oriented attitude', 'avoidance of reality and negative attitude', and 'short-term concern and absence of will to sustain'. Based on this, the analysis of board and lodging service graduates and middle surrender found that those who maintained the service for more than six months showed high keywords related to "will for employment and self-reliance," "self-management and future-oriented attitude," and those who gave up less than six months had many keywords related to "avoidance of reality and negative attitude" and "short-term concern and absence of will to sustain". These results confirm that there is a clear difference between the two groups.

With the continuous advancement of the data science and engineering, the combination of data analysis with other applications has been a trend. Data mining is in accordance with the established business goals from huge amounts of data to extract the effective potential and can be understood model of advanced treatment process. In this research paper, we conduct discussion on the execution of civil servants and professional ethics based on data analysis. Starting from the concept of data mining, this paper introduces the objects as the functions of data mining and mining process, mining algorithm combined with several kinds of common data mining, decision tree method, association rules method and neural network to its main ideas and improve the related description. We firstly review traditional data classification algorithms such as the Bayes classification algorithm to serve as the foundation. Later, we propose our method by using the modified SVM and deep neural network optimization. Then, we review the principles of civil servants and professional ethics with the combination of the prior discussion. The experimental result illustrates the feasibility and effectiveness of our method. We also discuss the future research plan in the final part.

With the rapid development of computer science and technology, data mining modelling techniques have emerged and rapidly developed as an alternative powerful meta-learning tool to accurately and fast analyze the massive volume of data generated by modern applications. The combination of data analysis technique and evaluation of public servant execution is urgently needed. Improve the execution of public servants at the grass-roots level is one of the important link to strengthen the construction of authority administrative efficiency of administrative goals is very important. Enhance the execution must first cultivate advanced concept, armed with advanced execution concept to the vast number of public servants at the grass-roots level. The assessment of public execution has a lot of traditional methods and models can be used but there is limitation. The limitation could be concluded as the following. Carelessness or poor sensitivity, At the grassroots level, the implementation of the main body of the general public servants at the grass-roots level and they can perform in place, one of the important factor is whether the leader on the work division of labor, organization, management and supervision effectively. In this paper, we conduct research on evaluation of public servant execution based on data mining technique and joint modeling analysis of multiple factors under big data environment. Firstly, we introduce some state-of-the-art clustering algorithm to serve as the basis of our model. Combined with deep neural network and optimization modelling, we propose our support vector machine based data clustering algorithm through multiple factor modelling. Subsequently, we discuss the principles on evaluation of public servant execution and process management. In the experimental part, we conduct experiment on both data clustering based data pre-processing step and the evaluation of elements’ weight for process management. The result indicates the most important factor for management and the feasibility and effectiveness of our proposed clustering method. Future potential research areas are also discussed in the final Section.

The development of the Internet brought us into an era of big data information, give people bring convenient while and also make people ragged when choosing the required information and recommendation system arises at the historic moment, and get the wide attention and applications. Therefore, to enhance the traditional method, we propose a novel intelligent recommendation algorithm based on Web data mining technique under the background of the deep neural network. Firstly, we review the state-of-the-art web data mining algorithms and revise the traditional ones with the parallel data mining algorithm on the multiple processors to perform tasks that will enhance the accuracy and efficiency. Then, we analyze basic neural network model through the inner connection and weight transfer. Later, we introduce the deep network structure to enhance the traditional network. Finally, we combine the revised prior theories into the recommending tasks for enhancement. The experimental analysis show that our algorithm accuracy is enhanced by the extent of 56% and overall time is reduced to the 87% of traditional ones which proves the feasibility. Later, more optimization work will be introduced to modify the current methodology.

In this paper, we present a sequence analysis method, which is one of the advanced data mining techniques, to identify and extract unique patterns from wafer manufacturing data. Wafer fabrication in the semiconductor industry is one of the most complex manufacturing processes. For such highly complicated operations, maintaining high yields through the statistical process control as a sole monitoring method for quality control is obviously inefficient. We thus investigate the intelligent and semi-automatic technique to help industrial engineers analyzing their production data. Our proposed method has the ability to induce patterns that can reveal and differentiate low performance processes from the normal ones. We also provide program coding of the proposed sequence analysis method, implemented with the R language, for easy experimental repetition.

Mining data in educational filed is an important and useful task for anybody related to educational institute. The useful information mined from the data can help with performance, guidance, teaching, planning and etc. for staff, students and instructors. Different educational data mining researches has been carried out on student data which includes the student’s basic achievements and educational background, academic scores and the amount of credit hours, but the relations between these are limited. Therefore in this paper we have described a system using data mining technologies such as decision tree. We have analyzed ID3 and J48 (C4.5) algorithms for predicting the scholarship winning chances on student data by translating the decision tree into “IF-THEN” rules and implementing these rules for prediction in our system called scholarship calculator. We have mined student data to calculate the chances of winning scholarship depending on their semester grades, position/rank of student in class, achievements, maximum and minimum amount of taken and allowed credit hours and extra curriculum activities. We found that ID3 works better even though J48 is faster in classifying data and creates a smaller tree than ID3. Because of ID3’s bigger tree it has more rule, more rules means more crosschecking and deeper decision, that’s why the predicted result was more accurate than J48. We have also described how our scholarship calculator works to predict and calculate the chances of winning scholarship. Performance evaluation is done and the results are also compared with already existing datasets. The developed system could be very useful in predicting student’s chances of winning scholarship from the first semester. It can help students to pinpoint the weak areas, which can be perfected with proper guidance from instructors and staff for better chances of winning scholarship.

In this age of information, which is available in different forms and these information play imperative part to enhance our knowledge base and so our social life. Information presentation is as important as information itself because to interpret the inside knowledge and interaction with it makes it more effective. This study investigates legend navigation interactive technique for comparative exploration of descriptive data mining results in order to communicate the insight information quickly, easily and in well understandable form. The legend navigation interactive mechanism is applied to two visualization techniques (column charts and bar charts) by performing descriptive data mining task on published Amazon dataset. The experimentation is done with 41 volunteers, selected by simple random sampling technique. The interactive technique is comprehensively analyzed in both visualization techniques considering visualization features. Equate the results with drill down interactive mechanism and discussed the utility of mechanism based on visualization features. The legend navigation approach and drill down interactive mechanism showed better performance in column chart comparatively.

Information technology plays vital role to enhance our knowledge and improve social life. Information presentation is as important as information itself, and interaction with these information enable one to understand these information quickly and easily. In this article, the information is explore up to one level granularity by introducing single level drill down interactive technique for descriptive data mining tasks results in order to convey inside of the data quickly, easily and effectively. The experimentation being done on Amazon dataset and two information visualization techniques i.e. column charts and bar charts. The interactive technique is comprehensively analyzing in both visualization techniques with respect to the visualization features. The drill down approach in column chart shows better performance comparatively.

In this paper, we introduce data mining techniques for short-term load forecasting (STLF). First, we use the K-mean algorithm to classify historical load data by season into four patterns. Second, we use the k-NN algorithm to divide the classified data into four patterns for Mondays, other weekdays, Saturdays, and Sundays. The classified data are used to develop a time series forecasting model. We then forecast the hourly load on weekdays and weekends, excluding special holidays. The historical load data are used as inputs for load forecasting. We compare our results with the KEPCO hourly record for 2008 and conclude that our approach is effective.