| CPC G06F 16/24578 (2019.01) [G06F 16/285 (2019.01)] | 20 Claims |
|
1. A processor-implemented method of analysing a data set, wherein the data set includes data for a plurality of variables resulting from a study, the method comprising:
uploading the data set by a user;
selecting via a user interface configured to facilitate such selection, by the user, one or more significant user-selected variables from the plurality of variables in the data set based on what interests the user and represents a priority & focus of a study, thereby ignoring remaining non-user-selected variables of the plurality of variables in the data set;
automatically selecting based on the type of the user-selected variable, by a processor, an appropriate statistical technique from a predefined library of statistical techniques for performing a correlation analysis for: (i) only measuring an interdependence of one or more user-selected variables associated with the data set and not remaining non-user-selected variables and (ii) only assessing a magnitude of the relationship between the one or more user-selected variables and not remaining non- user-selected variables;
performing, by the processor, a correlation for generating one or more correlation results comprising at least one of: a first set of user-selected variables related to the one or more significant user-selected variables, and a second set of user-selected variables related to the first set of user-selected variables, wherein the calculation results are determined independently of one another;
wherein performing the correlation comprises ranking the observations in an order of relevance by: classifying one or more variable observations associated with the data set based on at least one of: significance to the user, the variable type, the observation nature, and the observation result; generating a list of observations and insights based on the classification; running a relate process with two or more 1st, 2nd, and 3rd degree variables and collecting resulting observations; and compiling the observations across the data set for the two or more 1st, 2nd, and 3rd degree variables and ranking the observations by assigning a score to each observation;
and returning, by the processor, a list of key analysis, to the user, based on the one or more significant user-selected variables and the one or more correlation results related thereto, whereby the list of key analysis provides an insight to the user about the data set based only on the one or more user-selected variables.
|