Composite relationship discovery framework

1. A system comprising:

a memory storing processor-executable program code; and

a processing unit to execute the processor-executable program code to cause the system to:

receive a set of data including a plurality of continuous features and a discrete feature, each of the plurality of continuous features associated with a plurality of values and the discrete feature associated with a plurality of discrete values;

for each of the plurality of continuous features, determine a relationship factor representing a relationship between the discrete feature and the continuous feature based on the plurality of values associated with the continuous feature and the plurality of discrete values;

identify one of the plurality of continuous features associated with a largest one of the determined relationship factors;

for each of the plurality of continuous features other than the identified continuous feature, generate a correlation factor representing a correlation between the continuous feature and the identified continuous feature;

for each of the plurality of continuous features other than the identified continuous feature, determine a composite relationship score based on the relationship factor and the correlation factor associated with the feature; and

present a visualization associated with the discrete feature, the identified continuous feature, and a continuous feature associated with a largest composite relationship score.

2. A system according to claim 1 , wherein determination of a relationship factor for a continuous feature comprises:

derivation of ancillary statistics for each discrete value of the discrete feature based on the plurality of values associated with the continuous feature;

for each discrete value, derivation of primary metrics based on the ancillary metrics derived for the discrete value;

for each continuous feature, determination of a base relationship factor based on the primary metrics derived for the discrete value; and

for each continuous feature, determination of the relationship factor based on the base relationship factor determined for the continuous feature and a cardinality of the discrete feature.

3. A system according to claim 2 , wherein determination of the ancillary statistics for a discrete value comprises:

determination, for each continuous measure, of a mean of the values of the continuous measure associated with the discrete value; and

determination, for each continuous measure, of a variance of the values of the continuous measure associated with the discrete value.

4. A system according to claim 3 , wherein determination of the primary statistics for a discrete value comprises:

determination of a Sum of Squares Residual based on the determined variances; and

determination of a Sum of Squares Total based on the determined means.

5. A system according to claim 1 , wherein the presented visualization is associated with the discrete feature, the identified continuous feature, the continuous feature associated with the largest composite relationship score, and a continuous feature associated with a second-largest composite relationship score.

6. A method comprising:

receiving a set of data including a plurality of continuous features and a discrete feature, each of the plurality of continuous features associated with a plurality of values and the discrete feature associated with a plurality of discrete values;

for each of the plurality of continuous features, determining a relationship factor representing a relationship between the discrete feature and the continuous feature based on the plurality of values associated with the continuous feature and the plurality of discrete values;

identifying one of the plurality of continuous features associated with a largest one of the determined relationship factors;

for each of the plurality of continuous features other than the identified continuous feature, generating a correlation factor representing a correlation between the continuous feature and the identified continuous feature;

for each of the plurality of continuous features other than the identified continuous feature, determining a composite relationship score based on the relationship factor and the correlation factor associated with the feature; and

presenting a visualization associated with the discrete feature, the identified continuous feature, and a continuous feature associated with a largest composite relationship score.

7. A method according to claim 6 , wherein determining a relationship factor for a continuous feature comprises:

deriving ancillary statistics for each discrete value of the discrete feature based on the plurality of values associated with the continuous feature;

for each discrete value, deriving primary metrics based on the ancillary metrics derived for the discrete value;

for each continuous feature, determining a base relationship factor based on the primary metrics derived for the discrete value; and

for each continuous feature, determining the relationship factor based on the base relationship factor determined for the continuous feature and a cardinality of the discrete feature.

8. A method according to claim 7 , wherein determining the ancillary statistics for a discrete value comprises:

determining, for each continuous measure, of a mean of the values of the continuous measure associated with the discrete value; and

determining, for each continuous measure, of a variance of the values of the continuous measure associated with the discrete value.

9. A method according to claim 8 , wherein determining the primary statistics for a discrete value comprises:

determining a Sum of Squares Residual based on the determined variances; and

determining a Sum of Squares Total based on the determined means.

10. A method according to claim 6 , wherein the presented visualization is associated with the discrete feature, the identified continuous feature, the continuous feature associated with the largest composite relationship score, and a continuous feature associated with a second-largest composite relationship score.

11. A non-transitory medium storing processor-executable program code executable by a processing unit of a computing system to cause the computing system to:

receive a set of data including a plurality of continuous features and a discrete feature, each of the plurality of continuous features associated with a plurality of values and the discrete feature associated with a plurality of discrete values;

for each of the plurality of continuous features, determine a relationship factor representing a relationship between the discrete feature and the continuous feature based on the plurality of values associated with the continuous feature and the plurality of discrete values;

identify one of the plurality of continuous features associated with a largest one of the determined relationship factors;

for each of the plurality of continuous features other than the identified continuous feature, generate a correlation factor representing a correlation between the continuous feature and the identified continuous feature;

for each of the plurality of continuous features other than the identified continuous feature, determine a composite relationship score based on the relationship factor and the correlation factor associated with the feature; and

present a visualization associated with the discrete feature, the identified continuous feature, and a continuous feature associated with a largest composite relationship score.

12. A medium according to claim 11 , wherein determination of a relationship factor for a continuous feature comprises:

derivation of ancillary statistics for each discrete value of the discrete feature based on the plurality of values associated with the continuous feature;

for each discrete value, derivation of primary metrics based on the ancillary metrics derived for the discrete value;

for each continuous feature, determination of a base relationship factor based on the primary metrics derived for the discrete value; and

for each continuous feature, determination of the relationship factor based on the base relationship factor determined for the continuous feature and a cardinality of the discrete feature.

13. A medium according to claim 12 , wherein determination of the ancillary statistics for a discrete value comprises:

determination, for each continuous measure, of a mean of the values of the continuous measure associated with the discrete value; and

determination, for each continuous measure, of a variance of the values of the continuous measure associated with the discrete value.

14. A medium according to claim 13 , wherein determination of the primary statistics for a discrete value comprises:

determination of a Sum of Squares Residual based on the determined variances; and

determination of a Sum of Squares Total based on the determined means.

15. A medium according to claim 11 , wherein the presented visualization is associated with the discrete feature, the identified continuous feature, the continuous feature associated with the largest composite relationship score, and a continuous feature associated with a second-largest composite relationship score.