The contemporary society is shifting to a paradigm that creates valuable data by utilizing various types of data. The value of such data is highly appreciated, and the vast asset serves as useful analysis materials in information warfare. Entering the 21st century, the amount of data is increasing geometrically. The more data collected from such big data, the higher accuracy of its analysis, and the more efficient data extracted. Big data analysis consists of structured and unstructured data analysis. One of the analysis methods of text-type unstructured data is opinion mining. The analytical method is advanced text mining, which determines the status of the given sentence as either positive or negative based on sentiment words to analyze reputation. This study draws out an integrated Korean-based sentiment dictionary algorithm to enhance the accuracy of reputation analysis when using Korean as sentiment words in opinion mining analysis. Since Korean grammar shows different characteristics in vocabulary formation compared to English, SWN used in English grammar analysis has limitations in its application to Korean. Thus this study aims to build up a Korean-style sentiment dictionary, thereby proposing a methodology to increase the precision of opinion mining analysis.