| Sonstiges: |
- Nachgewiesen in: USPTO Patent Grants
- Sprachen: English
- Patent Number: 11868,217
- Publication Date: January 09, 2024
- Appl. No: 17/972556
- Application Filed: October 24, 2022
- Assignees: CDW LLC (Vernon Hills, IL, US)
- Claim: 1. A computer-implemented method for identifying a key of a database, the method comprising: for each column c1 in a first table, for each column c2 in a second table, computing a column similarity score by comparing a data type of the column c1 and a data type of the column c2, for each row value r1 in the first table, for each row value r2 in the second table, generating a value similarity score comparing the row value r1 with the row value r2, and when the column similarity score and/or the value similarity score exceed a respective threshold, identifying the column c2 in the second table as a foreign key of the column c1 in the first table.
- Claim: 2. The computer-implemented method of claim 1 , wherein the database key is a foreign key or a primary key.
- Claim: 3. The computer-implemented method of claim 2 wherein the foreign key is an intentional foreign key or a natural foreign key.
- Claim: 4. The computer-implemented method of claim 1 , wherein the database is a relational database.
- Claim: 5. The computer-implemented method of claim 1 , further comprising: determining one or more explicit key indications by analyzing a hints file.
- Claim: 6. The computer-implemented method of claim 1 , wherein computing the column similarity score by comparing the data type of the column c1 and the column c2 includes analyzing the data type of the column c1 and the column c2 using a trained machine learning model.
- Claim: 7. The computer-implemented method of claim 1 , wherein one or both of (i) the first table, and (ii) the second table include unstructured tabular data.
- Claim: 8. The computer-implemented method of claim 1 , wherein the database corresponds to a spreadsheet file and/or comma-separated value data.
- Claim: 9. The computer-implemented method of claim 1 , further comprising: analyzing one or both of (i) an existing indexes of the first table, and (ii) an existing index of the second table.
- Claim: 10. A computing system for identifying a key of a database, comprising: one or more processors; and one or more memories having stored thereon computer-executable instructions that, when executed by the one or more processors, cause the computing system to: for each column c1 in a first table, for each column c2 in a second table, compute a column similarity score by comparing a data type of the column c1 and a data type of the column c2, for each row value r1 in the first table, for each row value r2 in the second table, generate a value similarity score comparing the row value r1 with the row value r2, and when the column similarity score and/or the value similarity score exceed a respective threshold, identify the column c2 in the second table as a foreign key of the column c1 in the first table.
- Claim: 11. The computing system of claim 10 , wherein the database key is a foreign key or a primary key.
- Claim: 12. The computing system of claim 11 wherein the foreign key is an intentional foreign key or a natural foreign key.
- Claim: 13. The computing system of claim 10 , wherein the database is a relational database.
- Claim: 14. The computing system of claim 10 , the one or more memories having stored thereon computer-executable instructions that, when executed, cause the computing system to: determine one or more explicit key indications by analyzing a hints file.
- Claim: 15. The computing system of claim 10 , the one or more memories having stored thereon computer-executable instructions that, when executed, cause the computing system to: analyze the data type of the column c1 and the column c2 using a trained machine learning model.
- Claim: 16. The computing system of claim 10 , wherein one or both of (i) the first table, and (ii) the second table include unstructured tabular data.
- Claim: 17. The computing system of claim 10 , wherein the database corresponds to a spreadsheet file and/or comma-separated value data.
- Claim: 18. The computing system of claim 10 , the one or more memories having stored thereon computer-executable instructions that, when executed, cause the computing system to: analyze one or both of (i) an existing indexes of the first table, and (ii) an existing index of the second table.
- Claim: 19. A computer-readable medium having stored thereon computer-executable instructions that, when executed by one or more processors, cause a computer to: for each column c1 in a first table, for each column c2 in a second table, compute a column similarity score by comparing a data type of the column c1 and a data type of the column c2, for each row value r1 in the first table, for each row value r2 in the second table, generate a value similarity score comparing the row value r1 with the row value r2, and when the column similarity score and/or the value similarity score exceed a respective threshold, identify the column c2 in the second table as a foreign key of the column c1 in the first table.
- Claim: 20. The computer-readable medium of claim 19 , having stored thereon computer-executable instructions that, when executed by one or more processors, cause a computer to: analyze the data type of the column c1 and the column c2 using a trained machine learning model.
- Patent References Cited: 8386529 February 2013 Chaudhuri ; 9207930 December 2015 Srivas et al. ; 20110208748 August 2011 Chaudhuri ; 20160092494 March 2016 Kabra ; 20160371275 December 2016 Bernstein ; 20170169072 June 2017 Crawford ; 20180018579 January 2018 Xu ; 20180268050 September 2018 Kotorov ; 20190129959 May 2019 Jagwani ; 20190294621 September 2019 Konik ; 20200394542 December 2020 Buesser ; 20210149896 May 2021 Yu ; 20220147500 May 2022 Ellis
- Other References:
- Primary Examiner: Peng, Huawen A
- Attorney, Agent or Firm: MARSHALL, GERSTEIN & BORUN LLP ; Rueth, Randall G.
|
