經2017農地盤查結果可知全台超過13萬家違章工廠坐落於農地，儘管相關輔導政策持續推動，政策方向包含加強取締、劃設專區輔導合法等，仍未見成效，違章工廠持續出現於農地，顯示目前政策的實施方式需要透過更多討論與調整，並加強輔導政策的可行性。過去多數工廠區位相關研究皆以問卷方式取得廠商對於之看法，在大數據分析的時代，政府資訊逐漸公開，本研究跳脫傳統研究方法，以大量位在農地之違章工廠資料進行空間數據分析，檢視違章工廠的空間樣態及規模，並探究工廠與周邊區位特性資源之關係。
本研究以工業區位理論與工廠區位相關研究蒐集影響區位特性的資源變數，並以ArcGIS地理資訊系統運算彰化縣違章工廠與周邊資源之空間數據，其數據透過二元羅吉斯迴歸統計方法對影響違章工廠之變數進行解釋與預測。經研究發現，違章工廠區位特性相對依賴小地區資源，而後是大區域整體環境資源，另外也透過迴歸對違章工廠區位進行預測，希冀對違章工廠區位特性有進一步的了解，並可提供違章工廠輔導政策參考，尤其在輔導遷廠區位選擇與預防違章工廠發生上能更有序的執行。
According to the results of the 2017 Agricultural Land Inventory, There are more than 130,000 illegal factories in Taiwan are located in agricultural land. Although the relevant guidance policies continue to be promoted, the policy direction includes strengthening the ban and setting up special area guidance, and it still has no significant effect. The illegal factories continues to appear on agricultural lands. At present, the implementation of policies needs more discussions and adjustments to strengthen the feasibility of guidance policies.
In the past, most of the factory location related researches have obtained the opinions of the manufacturers by means of questionnaires. In the era of big data analysis, the government information is gradually open. This research breaks away from the traditional research methods and conducts spatial data analysis with a large number of illegal factory data in agricultural land. Examine the spatial form and scale of the illegal factory and explore the relationship between the factory and the surrounding location.
This study attempts to collect resource variables that affect the location characteristics with industrial location theory and factory location related research, and uses ArcGIS geographic information system to calculate the spatial data of illegal factories and surrounding resources in Changhua County. The data are interpreted and predicted by Binary Logistic regression statistics. The study found that the location characteristics of illegal factories are more dependent on small regional resources, and then the overall environmental resources of large areas. In addition, through the regression to predict the location of illegal factories, we hopes that there will be a better understanding of the location characteristics of illegal factories. This study could be for policy reference, especially in the guidance of relocation of the location and prevention of violations of the factory can be more orderly implementation.
