Modeling of food security based on the characteristics of the area will be affected by the geographical location which means that geographical location will affect the region's potential. Therefore, we need a method of statistical modeling that takes into account the geographical location or the location factor observations. In this case, the research variables could be global means that the location affects the response variables significantly; when some of the predictor variables are global and the other variables are local, then Geographically Weighted Ordinal Logistic Regression Semiparametric (GWOLRS) could be used to analyze the data. The data used is the resilience and food insecurity data in 2011 in East Java Province. The result shows that three predictor variables that influenced by the location are the percentage of poor (%), rice production per district (tons) and life expectancy (%). Those three predictor variables are local because they have significant influence in some districts/cities but had no significant effect in other districts/cities, while other two variables that are clean water and good quality road length (km) are assumed global because it is not a significant factor for the whole districts/towns in East Java.