Applying Data Mining Techniques to Customer Management of Children Protective Inoculation

• Main Authors: Hsiao-Yi Huang, 黃曉怡
• Other Authors: Der-Ming Liou
• Format: Others
• Language: zh-TW
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/96619484726249689921

碩士 === 國立陽明大學 === 衛生資訊與決策研究所 === 95 === People’s awareness about medical service has been raised during recent years, and service providers gradually can’t handle the market development. With the running of global budget , the development of medical service income is directly restricted. Besides, the management cost gradually increased, making exterior and interior environment less beneficial to hospital management. To make progress in the medical service market, we have to face the people’s demand and be more careful about investment of resources. Because national public health policy does a lot in protective affairs, the medical service trend is toward health promotion. This study is about vaccine application, according to customer relationship management strategy, to establish customer paying model with data mining tool and detect customer character, and hope this model could find out the group of customer who give the most and are most royal to the medical service institute. Besides, this study also discuss about the relationship between different vaccines, to understand the direction of self-paying vaccine program. Finally, with the coordination of cell flow, making the goal of employee and institute equal, and propose the proper business strategy, to promote self-paying vaccine, elevate the effect of hospital management. The study collected the neonatal vaccination record from birth to 4-year old in a medical center. Include the data of their mothers’ age, nationality, living areas. We set RFM (Rencency, Frequency, Money)classification model by the degree of customer giving and royalty, and apply the data in clustering mining results, find the specific group and the options that affect the classification by decision tree analysis. With the vaccination record of neonatal in the year of 2005, we discuss the regular and self-paying vaccine relationship through association mining. The result suggests that the royal customers found by RFM model, their characters are older mother age, higher proportion to take self-paying clinics and vaccines, live in the nearby areas. That shows there’s a number of customers could trust the medical center’s medical ability. There’s no specific demand on the physician, which means that the age of physician doesn’t affect the hospital profit. Mothers of foreign nations are not the high-profit groups. The meaningful factors include the visit numbder to the clinic, nationality of mothers, average clinic income, different physicians, and self-paying clinic ratio. There’s no correlation in regular and self-paying vaccines, not as expected, and mostly are replaceable modified hybrid vaccines to 3-in-1 vaccines. There’s correlation between streptococcal pneumonia vaccine and self-paying hepatitis vaccine, which could be a reference to clinical practice. Medical center’s pediatric clinics are in trouble because of inconvenience to clinic and higher health insurance fee, but this study shows that the royal customers to the medical center are with better economical ability, and it’s benefit to promote self-paying vaccines. The customer character is different from the local clinics. How to strengthen the hospital’s unique professional ability is the must-know direction. This study discussed about promotion strategy to royal customers. If further vaccination data could be collected and introduced to Data warehouse and dynametic update database, we could predict the shopping model of vaccination more precisely. Keyword:Customer Relational Management, Data Mining, Protective Inoculation