System and method for automated analysis of data to populate natural language description of data relationships

1. A computer implemented method of analysing a set of data records, wherein each data record comprises a value for each of a plurality of variables, the method comprising:

using one or more computer processors:

receiving a selection of a variable of interest from the plurality of variables;

analysing values in the data records of at least one variable of the plurality of variables other than the variable of interest by determining whether the at least one variable is correlated with the variable of interest and for each variable determined to be correlated with the variable of interest, determining the strength of the correlation;

using the analysis to create a statistical description of said at least one variable;

for one or more rules corresponding to the at least one variable, applying the rule to the plurality of variables;

for each rule having conditions of the rule met by the plurality of variables, choosing a template corresponding to the rule;

using the statistical description to populate the one or more templates corresponding to a rule, the templates populated for one or more variables other than the variable of interest, each template including a natural language description of a relationship between the variable of interest, or a value or range of values of the variable of interest, and the variable to which the statistical description applies; and

displaying the one or more populated templates including natural language descriptions for one or more of the variables, together with information relating to the strength of the correlation.

2. The method of claim 1 wherein said analysing is carried out on multiple variables other than the variable of interest and wherein at least one said template natural language description is populated for respective ones of said multiple variables.

3. The method of claim 1 wherein the populated natural language descriptions are presented as a list ranked by strength of correlation.

4. The method of claim 1 wherein said analysing is carried out on multiple variables other than the variable of interest and wherein multiple rules are available for at least one of the multiple variables.

5. The method of claim 1 in which each template comprises a natural language description and one or more placeholders, each placeholder to be populated with names of one or more of the plurality of variables and values of one or more of the plurality of variables.

6. The method of claim 1 comprising allocating the data records to respective groups, each group corresponding to a different value or range of values of the variable of interest, wherein the creation of a statistical description is performed for at least one of the respective groups.

7. The method of claim 1 comprising allocating the data records to respective groups, each group corresponding to a different value or range of values of said variable other than the variable of interest, wherein the creation of a statistical description is performed for at least one of the respective groups.

8. A system for the analysis of data records, each data record comprising a value for a plurality of variables, the system comprising:

one or more processors configured to:

allocate the data records to respective groups, each group corresponding to a different value or range or values of a variable of interest;

analyse the data records and determine whether any of the plurality of variables other than the variable of interest is correlated with the variable of interest;

determine the strength of correlation for each variable determined to be correlated with the variable of interest; and

for at least one group, create a statistical description of at least one variable of the plurality of variables other than the variable of interest;

for one or more rules corresponding to the at least one variable, applying the rule to the plurality of variables;

for each rule having conditions of the rule met by the plurality of variables, choosing a template corresponding to the rule; and

use the statistical description to populate the one or more templates corresponding to a rule, the templates populated for one or more variables other than the variable of interest, each template including a natural language description of a relationship between the variable of interest and the variable to which the statistical description applies; and

display the one or more populated templates including natural language descriptions for one or more of the variables used to populate the templates, together with information relating to the strength of the correlation.

9. The system of claim 8 wherein the one or more processors are configured to

carry out said creation of a statistical description for one or more variables which are correlated with the variable of interest.

10. The system of claim 8 in which each template comprises a natural language description and one or more placeholders, each placeholder to be populated with names of one or more of the plurality of variables and values of one or more of the plurality of variables.

11. The system of claim 8 further comprising an interface configured to receive a selection of a variable of interest from the plurality of variables.