A System of Visualizing Business Chart Using Text Mining Techniques

The selection of business charts based on data attributes and the purpose of visualization is an important consideration in improving accessibility of data interpretation and decision-making. However, the data visualization tools currently in usage are limited to the provision of numerous charts, and the selection of semantical-sensitive charts still falls back on personal experience and subjectivity. Therefore, the purpose of this study is to recommend proposed methods for improving the accuracy and efficiency of business charts in data usage by developing rules in accordance with data attributes and the purpose of visualization without specific knowledge required on the usage of business charts. For achieving the purpose, the definition and use cases of business charts were collected through web crawling, and keywords characterized each business chart were extracted through TF -IDF analysis. Subsequently, keywords matrix was created according to the associations of these extracted keywords with each business chart and used as recommendation rules.