Applying Fuzzy Ontology to develop Race Forecast System

• Main Authors: Yu-Shien Chen, 陳昱先
• Other Authors: Yu-Liang Chi
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/87317564239841263312

碩士 === 中原大學 === 資訊管理研究所 === 101 === Since ancient time, forecasting has seen everywhere in human’s history. People used forecasting to help decision-making, moreover, in the history of the IT development, many systems using historical data sets to train computer for assist the human decision-making, but in the past, training sets with the information was focused on the crisp information set, therefore, the system couldn’t fully expressed human’s cognition and feelings, which cause imprecisely decision. Be able to enhance the accuracy of the forecasting and break the traditional training way of crisp set, this research tries to use fuzzy theory combined ontology with specific areas of needs analysis, collecting the relevant domain knowledge, then build domain knowledge completely. Otherwise, machine didn’t recognize human’s experience and cognitive knowledge, so the research defines fuzzy concepts and membership function by fuzzy theory for fuzzy factors in knowledge domain, thus machine can know these data with number interval. The user can query through the Web-based interface, directly access to the forecast results. The system using Major League Baseball for the case of race prediction, main system design includes (1) knowledge acquisition and knowledge modeling; (2) fuzzy theory combination; (3) forecasting system design, which enhance the accuracy of the prediction system. After the experiment with MLB race stat, fuzzy factors correct forecasting outcomes.