| Summary: | 碩士 === 國立成功大學 === 高階管理碩士在職專班(EMBA) === 104 === Taiwan government organizations have implemented Open Government Data (OGD) policies to make their data public available. Over ten thousand datasets are currently available in data.gov.tw within short period. The aims of this study were to examine the quality and content of health-related datasets in data.gov.tw. and to develop three data visualization models for health policy decision making. Until end of October 2015, there were 840 health-related datasets available in data.gov.tw. However, many organizations decomposed the datasets by year or by disease, which did not conform OGD principle of completeness. About one fifth of datasets in which the formats were not machine processable. With regard to the content of the datasets, disease statistics, quality indicators of medical care, lists commitee and items, budegets were the four main themes accounted for seventy percent of all datasets. The most popular downloaded datasets were basic information of pharmacies, inappropriate food information, cause of death statistics, name list of PIC/S GMP pharmaceutical industries and hospital beds statistics. With regard to the second objective, I used Tableau and Excel Power Pivot to develop three data visualization health policy decision making, i.e., problems identification, priority setting, and performace evaluation. The three models have been uploaded into Tableau Public and freely available to general public. Users of the three models could select the year, the cause of death, and the area they concerned to produce the statistical plot they want. In conclusion, health-related organization should conform the OGD principles to release the datasets as complete as possible. Further studies are needed to survey the applications of using health-realted OGD and explore the potential of integrating health-related OGD datasets with non-health OGD datasets. The three models developed in this study should be tested by stakeholds of health policies and some feedbacks to improve the performance of the models and in the long run promote the quality of health policy decision making.
|
|---|
