Data set discovery engine comprising relativistic retriever

1. An apparatus comprising:

a processing platform implementing a data set discovery engine;

wherein the data set discovery engine comprises:

a data set indexer configured to generate similarity indexes for a plurality of data sets each comprising a plurality of data items; and

a relativistic retriever coupled to the data set indexer and configured to obtain a suitability template for a query and to execute the query against one or more of the similarity indexes based at least in part on the suitability template;

the suitability template limiting usage of particular ones of the data sets to particular data set discovery contexts;

wherein a given one of the similarity indexes comprises, in addition to indexing information characterizing data items of its corresponding one or more data sets, at least first and second auxiliary information generated from respective ones of at least first and second different similarity measures of a plurality of different similarity measures supported by the data set discovery engine;

the first auxiliary information being computed from the data items of the corresponding one or more data sets using the first similarity measure;

the second auxiliary information being computed from the data items of the corresponding one or more data sets using the second similarity measure; and

wherein the processing platform comprises one or more processing devices each comprising a processor coupled to a memory.

2. The apparatus of claim 1 wherein the plurality of different similarity measures supported by the data set discovery engine comprise at least one frequency-based similarity measure and at least one non-frequency-based similarity measure, and wherein an arbitrary set of two or more similarity measures selected from the plurality of different similarity measures are utilized in conjunction with execution of the query.

3. The apparatus of claim 2 wherein said at least one frequency-based similarity measure comprises at least one of a term frequency-inverse document frequency (TF-IDF) similarity measure and a probabilistic similarity measure, and wherein said at least one non-frequency-based similarity measure comprises a Jaccard similarity measure.

4. The apparatus of claim 1 wherein the suitability template is part of the query.

5. The apparatus of claim 1 wherein at least one of the first and second similarity measures is automatically selected based at least in part on the suitability template.

6. The apparatus of claim 1 wherein functionality for generating at least one of the first and second similarity measures is provided to the data set discovery engine on a pluggable basis via an application programming interface.

7. The apparatus of claim 1 wherein the suitability template characterizes suitability for at least one of a particular purpose, a particular goal and a particular analytic role.

8. The apparatus of claim 1 wherein the suitability template is associated with at least one target data set and further wherein the data set indexer is configured to generate similarity indexes for a plurality of target data sets each associated with one or more suitability templates.

9. The apparatus of claim 1 wherein the first and second similarity measures have respective first and second weights assigned thereto and wherein said assigned weights are taken into account by the relativistic retriever in recommending in conjunction with execution of the query one or more of the plurality of data sets that exhibit a specified degree of similarity to at least one target data set associated with the suitability template.

10. The apparatus of claim 9 wherein at least one of the similarity measures and at least one of the weights assigned to the respective similarity measures are adjusted over time through machine learning based at least in part on user interaction with particular data sets recommended by the relativistic retriever, said user interaction comprising at least one of user selection of one or more of the recommended data sets and user evaluation of one or more of the recommended data sets.

11. The apparatus of claim 1 wherein the relativistic retriever is configured to retrieve particular ones of the data sets in conjunction with execution of the query and to order the retrieved data sets in terms of their respective degrees of similarity to a given target data set associated with the suitability template.

12. The apparatus of claim 1 wherein the relativistic retriever is configured to identify in conjunction with execution of the query one or more of:

(i) relationships among retrieved data sets; and

(ii) relationships among the relationships in (i).

13. The apparatus of claim 12 wherein the relativistic retriever is configured to permit a user to navigate the retrieved data sets in accordance with the identified relationships and one or more additional relationships, the one or more additional relationships comprising at least one relationship selected from:

(i) other relationships among the retrieved data sets and other data sets; and

(ii) other relationships that the other data sets participate in.

14. The apparatus of claim 12 wherein the identified relationships are determined at least in part based on one or more of:

(i) similarity of a target data set associated with the suitability template to one or more other target data sets associated with one or more other suitability templates; and

(ii) one or more semantic considerations derived from a semantic hierarchy of data sets and relationships.

15. The apparatus of claim 1 wherein the relativistic retriever is configured to propagate at least one of a tag and a classification from one of the data sets to another one of the data sets in conjunction with the execution of the query.

16. The apparatus of claim 15 wherein the suitability template is a source for at least a portion of at least one of the tag and the classification.

17. A method comprising:

generating similarity indexes for a plurality of data sets each comprising a plurality of data items;

obtaining a suitability template for a query; and

executing the query against one or more of the similarity indexes based at least in part on the suitability template;

the suitability template limiting usage of particular ones of the data sets to particular data set discovery contexts;

wherein a given one of the similarity indexes comprises, in addition to indexing information characterizing data items of its corresponding one or more data sets, at least first and second auxiliary information generated from respective ones of at least first and second different similarity measures of a plurality of different similarity measures;

the first auxiliary information being computed from the data items of the corresponding one or more data sets using the first similarity measure;

the second auxiliary information being computed from the data items of the corresponding one or more data sets using the second similarity measure; and

wherein the generating, obtaining and executing are performed by a processing platform comprising one or more processing devices.

18. The method of claim 17 further comprising:

retrieving particular ones of the data sets in conjunction with execution of the query;

identifying one or more of:

(i) relationships among the retrieved data sets; and

(ii) relationships among the relationships in (i); and

permitting a user to navigate the retrieved data sets in accordance with the identified relationships and one or more additional relationships.

19. A computer program product comprising a non-transitory processor-readable storage medium having one or more software programs embodied therein, wherein the one or more software programs when executed by at least one processing device of a processing platform cause the processing device:

to generate similarity indexes for a plurality of data sets each comprising a plurality of data items;

to obtain a suitability template for a query; and

to execute the query against one or more of the similarity indexes based at least in part on the suitability template;

the suitability template limiting usage of particular ones of the data sets to particular data set discovery contexts;

wherein a given one of the similarity indexes comprises, in addition to indexing information characterizing data items of its corresponding one or more data sets, at least first and second auxiliary information generated from respective ones of at least first and second different similarity measures of a plurality of different similarity measures;

the first auxiliary information being computed from the data items of the corresponding one or more data sets using the first similarity measure;

the second auxiliary information being computed from the data items of the corresponding one or more data sets using the second similarity measure.

20. The computer program product of claim 19 wherein the one or more software programs when executed by at least one processing device of the processing platform further cause the processing device:

to retrieve particular ones of the data sets in conjunction with execution of the query;

to identify one or more of:

(i) relationships among the retrieved data sets; and

(ii) relationships among the relationships in (i); and

to permit a user to navigate the retrieved data sets in accordance with the identified relationships and one or more additional relationships.