대용량 온톨로지 데이터 처리를 위한 효율적인 저장구조 설계 및 구현

성정환 창원대학교 대학원 2006 국내석사

Recently, many researchers are proceeding the study as to knowledge management, human resource management, information retrieval using the ontology technology. Hence it is necessary to study for the efficient ontology storage structure before using the ontology technology. This paper proposed the efficient ontology storage structure that was divided into the relation information and the data information for processing a large ontology data. For experiment of the proposed ontology storage structure in this paper, we applied a family ontology schema of Oracle 10g Spatial RDF. Experiment data consist of family members from 1st ancestor to 14th descendant. The experiment of this paper compared vertical storage model, property per table storage model, horizontal storage model, hybrid storage model with proposed storage structure. In the result of experiment, proposed storage structure is more effective than previous storage model in the storage area, and query performance is also improved.

개인중심 지식네트워크시스템의 설계 및 구현

강승준 창원대학교 2009 국내석사

This paper proposes a new type of Knowledge Network System to manage knowledge in a systematic manner and to encourage voluntary knowledge sharing through knowledge-to-people and people-to-people networks. The proposed system consists of Personal Knowledge Management Subsystem and Network-based Information Management Subsystem. The Personal Knowledge Management Subsystem consists of Personal Knowledge Map Management Module to systematically manage knowledge structures by individual, Secure Cabinet Module to protect knowledge files and Personal Page Management Module to manage personalized menus and information. The Network-based Information Management Subsystem comprises Knowledge Network Management Module which manages and shares knowledge among members and Collaboration Management Module which assists collaboration among members of an organization.

태그 온톨로지를 이용한 태그 추천 시스템

김재승 창원대학교 2009 국내석사

사용자 정의 자원이 증가함에 따라 자원에 의미를 부여하고 효율적 관리를 위해 태그를 부여하는 태깅 시스템과 사용자에게 자원에 대한 적합한 태그를 추천하는 태그 추천 시스템이 증가하고 있다. 태그는 자원의 의미와 사용자의 의도를 포함하고 있는 키워드이다. 태깅 시스템은 사용자들이 자원의 공유와 관리를 위해서 자원에 태그를 부여하도록 하는 시스템이며 태그 추천 시스템은 사용자에게 적합한 태그를 추천해 주는 시스템이다. 기존의 시스템은 자원에 대한 메타데이터 분석, 협업 태깅, 온톨로지 등을 통하여 사용자에게 의미 있는 태그를 추천하려 하였다. 하지만 기존의 시스템은 사용자가 작성한 키워드를 기반으로 태그를 추천하기 때문에 용어의 맞춤법 및 표준화에 문제가 있다. 최근에 이러한 문제를 해결하기 위한 연구가 증가하고 있다. 본 논문에서는 이러한 문제를 개선하기 위해서 온톨로지를 이용하여 태그를 표준화 하였으며 사용자에게 의미 있는 태그를 추천하기 위해서 TWCIDF 알고리즘과 TWCITC 알고리즘을 제안 하였다. TWCIDF 는 대량의 웹 문서에 대한 자동 태깅 기법을 위해서 이용되며 TWCITC 는 사용자가 웹 상에 등록하는 신규 문서에 대한 태그 추천 기법에 사용 된다. 두 가지 기법을 실험하기 위해서 하이브레인 넷의 ‘해외 연수’ 부분의 2008년 8월 한달 간의 830건 데이터를 이용하여 온톨로지와 불용어사전 등을 구축하였으며 2008년 7월과 8월 두달 간의 1440건의 데이터를 기반으로 실험하였다. In this paper, we propose a Tag Recommendation System using Tag ontology. This system is recommended standardized tag to user. Tag is a keyword to express about resources. However, it has many types, even having the same meaning. So, we are standardized a tag by using Tag Ontology. Therefore, we propose two algorithmes to recommend appropriate tag to user. TWCIDF recommend tag for unstadardization documents and TWCITC recommend tag for new doucment.

자동차 모듈 BOM 구성 및 제품구조 관리를 지원하는 PDM 시스템 개발

천동호 창원대학교 2005 국내석사

PDM(Product Data Management) is the system that collects and integrated manages of the various information related to research & development of the products centering around design field in the process of product production. This research is the newly developed PDM system which support automotive module BOM configuration & product structure management as using registered design information. In this research, with improving an existing method to compose and register BOM manually, the procedures to receive various design information through on-line from automotive companies, resister and process it inside of PDM are restructured in order to manage complex and various specification of automotive parts efficiently which are produced in the process of R&D for a new car. This research which the improved PDM system is applied makes shortening period, improving product quality, reducing cost in the product development as well as concurrent engineering possible and competitive power of company high.

지식 거래 서비스를 위한 의미 기반의 검색 기법

송성광 창원대학교 2009 국내석사

In this paper, we propose a Semantic-based Search Techniques that is able to search related tags, related products, related users using with defined rules to knowledge commercial services and rule-based search. Ontology Inferencing is not suited to search a number of individuals ontology or has problem that is not able deal with datatype property before this. Hence we propose this techniques that define rules to knowledge commercial services and use rule-based search to improvement it. This techniques can apply web2.0-based knowledge commercial services site using with searching related tags, related products and related users.

빅 데이터 분석 기법을 이용한 한국 프로야구 타자 평가 지표 개발

정진상 창원대학교 2014 국내석사

When the size of data is very big so that we cannot collect, manage and process that data with usual software program, such data is called a Big data. Baseball is very appropriate area to analyze using Big data. Because there are lots of things to research and investigate such as the direction of the ball hit by player, the movement of a fielder and the course of the ball pitched by player. From now on, departing from using usual indicators like a batting average, analyzing non-standardized data is going to be significant issue in baseball. This thesis suggests Highballpoint to estimate hitter in Korea Baseball League using Big data. This consists of 'Base score', 'Distribution score' and 'Pitch waste score'. The 'Base score' is a point of event for result of hitting calculated on 'Run value' based on Run. The 'Contribution score' is a point of context for result of hitting based on 'Win expectancy' which calculates probability of winning. The 'Pitch waste score' is a point for waiting a number of balls when pitcher pitches hitter. The Highballpoint considers all events calculated on Run and winning of team unlike the OPS and Casspoint which is another estimation of hitter. The OPS does not consider several events except single, double, triple, homerun to estimate hitter. The Casspoint highly estimates homerun score more than the others therefore hitter who hit homerun is ranked on high position. The Highballpoint calculates all events based on empirical data. It also estimates result of hitting considered team win or lose.

Social Network 구성을 위한 OWL 기반의 사용자 모델링 기법

전인하 창원대학교 대학원 2006 국내석사

In this paper, we proposed a user modeling technique based on ontology for building social network. User modeling technique is composed of user ontology creation process for expressing user information and social network construction process that is applied closeness weight. User ontology schema is defined by expanding and refering the candidate schema of HR-XML and is created by OWL(Web Ontology Language). We created ontology instances using user ontology schema and defined a user ontology rule using SWRL(Semantic Web Rule Language) for supporting a reasoning service. And we analyzed an information of association connection attribute based on user ontology instance for represent semantic relation between users. In this paper, we computed a closeness between users using the information of association accomplishment and extracted direct relation that applied a closeness weight. We composed the user social network with closeness weight by expanding the direct relation between users.

Web 2.0 기술을 이용한 SCORM 기반의 적응적 e-Learning 시스템 설계 및 구현

김영균 창원대학교 대학원 2006 국내석사

In this paper, we proposed a e-Learning system based on SCORM which supports adaptive learning to learner using Web2.0 technology. This system collect learner's learning behaviors based SCORM CMI data model using learning tracking module. This system store collected learning behaviour information as instance using proposed LBML schema. Learner behavior analyzing module evaluate the level of learner and analyze learned contents using LBML instance. This system supported contents with feedback information as compare rule defined by teacher and learner's learning behavior information. And this system supported adaptive learning contents using information that was analyzed by learning behaviour from a learner. Related contents can be recommended using TFIDF algorithm by text-mining technique. Learning Support Module is implemented taking feedback information by users as adaptive learning contents in realtime.

야구 경기에서 빅데이터 분석과 마르코프 연쇄를 이용한 득점 예측 모형

문형우 창원대학교 2014 국내박사

This paper presents a new model for predicting the number of runs scored in a baseball game on the basis of a big data analysis and a Markov chain. To this end, a database model was designed to implement a systematic management of the large amount of baseball game data. The MapReduce technique in the Hadoop framework, a method widely used in big data analyses, was used for effective storage and systematic management of the large amount of game score data consisting of unstructured text data. For efficient configuration of the proposed Markov chain-based score prediction model, the probabilities of advancing and hitting were redefined to accurately simulate the real-world baseball game situations. Using the probabilities of advancing and hitting, we obtained the score distributions and the number of batters for each inning, and constructed the Markov chain model to predict the scoring runs in each game. A -test was used for verifying the difference in the probabilities of advancing and hitting between right- and left-handed pitchers, and a score prediction model reflecting the characteristics of right- and left-handed pitchers was constructed. Real game data from korean professional baseball were used for experimentally proving the efficiency of the proposed prediction model. The experiment also included a score prediction model that takes into account the characteristics of left- or right-handed pitchers by reflecting their respective probabilities of advancing and hitting. The proposed model is expected to be useful in establishing strategies for deciding the batting order or enhancing game performance through efficient predictions of the scoring runs and the winning odds.

지식 거래 서비스를 위한 온톨로지 모델 설계 및 생성

문형우 창원대학교 2009 국내석사

Knowledge Commerce Service is a Service for intermediating knowledge product business that is a higer value-added like translation, consulting and design. The core of knowledge Commerce Service is managing knowledge product, intermediating knowledge business. This paper designs effective ontology model for Knowlege business service. This model is conceptual ontology organized user, product and tag ontologies to effective manage unstandardized knowledge product and relation ontologies that define relation between ontologies.