Data quality enhancement for smart grid applications
First Claim
Patent Images
1. A method comprising:
- retrieving data from an electricity smart grid;
encoding said data into a data modeling technique;
performing a data quality rules discovery process using a data set of said electricity smart grid and said data modeling technique encoded with said data to said data set to produce a plurality of data quality rules associated with said electricity smart grid, said data quality rules discovery process comprising;
applying each of a set of candidate conditional functional dependencies to a data segment having a predetermined length of data points of said data set; and
after said applying, if a candidate conditional functional dependency has a result signature that does not meet a predetermined expectation, refining that candidate conditional functional dependency by eliminating an attribute in that candidate conditional functional dependency that took on the most values throughout the data set.
3 Assignments
0 Petitions
Accused Products
Abstract
A method, in one embodiment, can include encoding knowledge about a topic domain into a data modeling technique. Additionally, a set of candidate conditional functional dependencies can be generated based on a data set of the topic domain. Moreover, the set of candidate conditional functional dependencies and the data modeling technique encoded with the topic domain knowledge can be applied to the data set to obtain a plurality of data quality rules for the data set.
17 Citations
20 Claims
-
1. A method comprising:
-
retrieving data from an electricity smart grid; encoding said data into a data modeling technique; performing a data quality rules discovery process using a data set of said electricity smart grid and said data modeling technique encoded with said data to said data set to produce a plurality of data quality rules associated with said electricity smart grid, said data quality rules discovery process comprising; applying each of a set of candidate conditional functional dependencies to a data segment having a predetermined length of data points of said data set; and after said applying, if a candidate conditional functional dependency has a result signature that does not meet a predetermined expectation, refining that candidate conditional functional dependency by eliminating an attribute in that candidate conditional functional dependency that took on the most values throughout the data set. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
-
-
9. A method comprising:
-
retrieving data about a health care domain; encoding said data into a data modeling technique; performing a data quality rules discovery process using a data set of said health care domain and said data modeling technique encoded with said data to said data set to produce a plurality of data quality rules for said data set associated with said health care domain, said data quality rules discovery process comprising; applying each of a set of candidate conditional functional dependencies to a data segment having a predetermined length of data points of said data set; and after said applying, if a candidate conditional functional dependency has a result signature that does not meet a predetermined expectation, refining that candidate conditional functional dependency by eliminating an attribute in that candidate conditional functional dependency that took on the most values throughout the data set; and performing an analysis on said plurality of data quality rules to generate a derivative data quality rule for said data set. - View Dependent Claims (10, 11, 12, 13, 14, 15, 16)
-
-
17. A system comprising:
-
an interface capable of receiving data about an electricity smart grid; a first module coupled to said interface, said first module configured to encode said data into a data modeling technique; and a discovery engine coupled to said first module, said discovery engine configured to perform a data quality rules discovery process using a data set of said electricity smart grid; and said data modeling technique encoded with said data to said data set to produce a plurality of data quality rules associated with said electricity smart grid, said data quality rules discovery process comprising; applying each of a set of candidate conditional functional dependencies to a data segment having a predetermined length of data points of said data set; and after said applying, if a candidate conditional functional dependency has a result signature that does not meet a predetermined expectation, refining that candidate conditional functional dependency by eliminating an attribute in that candidate conditional functional dependency that took on the most values throughout the data set. - View Dependent Claims (18, 19, 20)
-
Specification