Visual Data Analysis Based on Attribute Hierarchy

• Main Authors: Yi-Chen Shen, 沈義誠
• Other Authors: Chung-Chain Hsu
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/92945497919912335449

碩士 === 國立雲林科技大學 === 資訊管理研究所 === 87 === Useful patterns or knowledge may hide in huge amount of data. By properly analyzing and mining the data, an analyst can uncover and provide useful information to executives for decision making. In order to facilitate data exploration task to those data analysts who are not computer professionals and help user explore data from different angles and at multiple abstraction levels, a system has to be user-friendly. This study proposed a graphical data analysis model based on attribute hierarchy and an extended relational query language. The model is capable of supporting unstructured what-if type data analysis. Attribute hierarchies that store domain knowledge play an important role during data analysis sessions. Generalization can be performed according to attribute hierarchies for high level queries and analysis. The proposed model provides an interactive data analysis environment for data analysts who are not computer profes-sionals to perform drill-down, roll-up and summarization along any attribute dimensions and to perform what-if data analysis.