| 摘要 |
The present invention concerns a device and a method for pattern analysis of large volumes of data in order to make it possible to distinguish tendencies to co-variations and developments in time from one large database, preferably one composed by a number of smaller databases. The data in the database is divided into on the one hand objects and on the other variables. The method is characterized by first grouping the volume of data, whereby the variables are grouped to dimensions and objects to categories based on a variable or a dimension, by thereafter storing the aggregated database thus formed, by sorting the database with respect to dimensions and categories in order to position similar dimensions and categories in proximity to one another and by graphically presenting the volume of data, the grouping parameters, the sorting parameters, and the presenting parameters being controlled by the user, and by repeating the method the desired number of times while changing one or several of the parameters mentioned. |
