Zero-Inflated Patent Data Analysis Using Generating Synthetic Samples

Due to the expansion of internet, we encounter various types big data such as web documents or sensing data. Compared traditional small experimental samples, provide more chances find hidden and novel patterns with analysis using statistics machine learning algorithms. However, use increases, problems also occur. One them is a zero-inflated problem in structured preprocessed from Most count values are zeros because specific word found only some documents. In particular, since most patent form text document, they affected by problem. To solve this problem, propose generation synthetic samples statistical inference tree structure. Using document simulation data, verify performance validity our proposed method. paper, focus on keyword analysis, just like other