Apparatus and method for assessing relevant categories and measures for use in data analyses
View Patent ↗A computer readable storage medium includes executable instructions to receive a relevancy parameter. The relevancy parameter is searched in a category ensemble including a set of categories, where the category ensemble overlies a dataset. A first order of the set of categories is created based on the relevancy of the relevancy parameter to each category in the set of categories, where the relevancy is a degree of match to the relevancy parameter. A second order of the set of categories is created based on the merit of each category to partition the dataset. The first order and the second order are combined into a final order, which is returned. A measure in the category ensemble is searched based on the relevancy parameter. The measure is returned as a selected measure, where the selected measure is a codomain of a visualization depicting a portion of the dataset.
1. A computer readable storage medium, comprising executable instructions to:
receive a selection of a category ensemble overlying a dataset;
define an analysis view of the category ensemble, wherein the analysis view comprises:
a set of categories divided into:
a first subset of unselected categories, and
a second subset of selected categories,
a set of measures, and
a visualization for a subset of the dataset, including a preselected measure that is a highly relevant measure in the category ensemble, wherein the preselected measure is a codomain of the visualization;
receive a relevancy parameter;
conduct a search based on the relevancy parameter in the category ensemble;
create a first order of the first subset of unselected categories based on relevancy, wherein the first order based on the relevancy is based on relevancy of the relevancy parameter to each category in the set of categories;
create a second order of the first subset of unselected categories based on merit, wherein merit is a value which indicates efficiency for partitioning the dataset;
combine the first order and the second order into a final order for the first subset of unselected categories of the set of categories; and
update the analysis view by displaying the first subset of unselected categories in the final order.
2. The computer readable storage medium of claim 1 further comprising executable instructions to:
conduct a search based on the relevancy parameter over a plurality of category ensembles; and
order the plurality of category ensembles by relevancy.
3. The computer readable storage medium of claim 1 wherein a high value for merit is taken from the combination of two or more values selected from high entropy, high coverage and low cardinality.
4. The computer readable storage medium of claim 1 , wherein relevancy is based on at least one of keyword, usage metrics, data source information, data quality level, data lineage metadata, favorite information, data design information and data metadata.
5. The computer readable storage medium of claim 4 wherein relevancy based on keyword is selected from the group comprising equality of keyword to name, similarity of keyword to name, equality of keyword to metadata, and similarity of keyword to metadata.
6. The computer readable storage medium of claim 1 further comprising executable instructions to:
accept the selection of an attribute in a category in the set of categories, wherein the category includes a set of attributes; and
filter the dataset using the attribute as a filter value.
7. The computer readable storage medium of claim 6 further comprising executable instructions to display the second subset of selected categories in a third order, wherein the third order reflects the sequence in which the categories in the second subset of selected categories were selected.
8. The computer readable storage medium of claim 6 further comprising executable instructions to determine another final order for the first subset of unselected categories.
9. The computer readable storage medium of claim 1 further comprising executable instructions to:
select a highly relevant category in the category ensemble; and
return the category as a preselected domain in a visualization.
10. The computer readable storage medium of claim 1 wherein the final order is selected from the group comprising:
heterogeneous in terms of the first order and the second order of the first subset of unselected categories, and
homogenous in terms of the first order and the second order of the first subset of unselected categories.
11. The computer readable storage medium of claim 1 further comprising executable instructions to:
accept the selection of a category in the set of categories; and
update the analysis view to include the category as a breakdown category.
12. The computer readable storage medium of claim 1 further comprising executable instructions to:
display the analysis view in a graphical user interface, wherein:
the first subset of unselected categories is made available for selection,
the second subset of selected categories is made available for unselection,
the set of measures is made available for selection or unselection,
a selection change includes selection and unselection, and
receive a representative selection change created by a user's interaction with the graphical user interface; and
refresh the analysis view based on the representative selection change.
13. The computer readable storage medium of claim 1 , wherein:
each of the set of measures, the set of categories and the category ensemble is associated with two or more pieces of metadata selected from: name, multilingual name, description, user tags, user comments, usage metrics, data source information, data quality level, data lineage information, favorite information, data design information and data information.