DICTIONARY ENTRY NAME GENERATOR

• US 20130332454A1
• Filed: 06/06/2012
• Published: 12/12/2013
• Est. Priority Date: 06/06/2012
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP AG (SAP SE)
• Inventors: Stuhec, Gunther, Weissmann, Dirk
• Granted Patent: US 8,892,575 B2
• Time in Patent Office: Days
• US Class Current: 707/730
• CPC Class Codes: G06F 16/211 Schema design and management
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/242 Dictionaries

Abstract

A method for building dictionary entry names for data elements of a canonical data model includes identifying candidate terms for the dictionary entry name of a node or equivalence class of the canonical data model. The method includes counting a frequency of occurrence of candidate terms in use and based on the use counts creating a candidate ordering of terms for the complete ordered dictionary entry name of the node or equivalence class. The method further includes validating the candidate ordering of terms for the complete ordered dictionary entry name of the node or equivalence class by comparison of the ordering with reliable dictionary entry name entries in a database and/or by usage counts in search engine results.

Claims

1. A microprocessor-implemented method, comprising:

receiving input data on candidate terms for building dictionary entry names (DENs) for a canonical data model (CDM);

processing the input data to identify candidate terms for the DEN of a node or equivalence class of the CDM;

counting a frequency of occurrence of candidate terms in the input data;

creating a candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM based at least in part on the counted frequency of occurrence of candidate terms; and

validating the candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM by at least, one of comparison of the ordering with reliable DEN entries in a database and evaluation of relative usage counts in search engine results.

2. The method of claim 1, wherein receiving input data on candidate terms or names for building DENs for the CDM includes receiving information on equivalence classes of data elements that are used in different data structures, formats or models.

3. The method of claim 1, wherein receiving input data on candidate terms or names for building DENs for the CDM includes receiving names and natural language definitions of data elements used in different data structures, formats and models, and receiving corresponding data element, names and definitions used in UN/CEFACT Core Component Technical Specification (CCTS).

4. The method of claim 1, wherein processing the input data to identify candidate terms for the DEN for the node or equivalence class of the CDM includes resolving camel case words.

5. The method of claim 1, wherein processing the input data to identify candidate terms for the DEN for the node or equivalence class of the CDM includes part-of-speech tagging of terms as nouns, verbs or adjectives.

6. The method of claim 5, wherein processing the input data to identify candidate terms for the DEN for the node or equivalence class of the CDM further includes at least one of lemmatization, consolidation of related terms, spell checking and automatic correction of spell, errors, resolving abbreviations, and interpretation of compound nouns.

7. The method of claim 5, wherein processing the input data to identify candidate terms for the DEN for the node or equivalence class of the CDM further includes processing verbs and adjectives that are orthographically similar to verbs as nouns.

8. The method of claim 5, wherein processing the input data to identify candidate terms for the DEN for the node or equivalence class of the CDM further includes treating adjectives that are orthographically dissimilar to verbs as preferred DEN term qualifiers.

9. The method of claim 1, wherein counting the frequency of occurrence of candidate terms in the input data includes weighing the frequencies of occurrence of candidate terms by assigning relative weights to occurrences of candidate terms in definitions, names and DENs portions of the input data.

10. The method of claim 1, wherein creating the candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM includes creating an ordering of one or more terms having the highest frequencies of occurrence.

11. The method of claim 1, wherein creating the candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM includes inheriting a first term from a root node.

12. The method of claim 1, wherein creating the candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM includes determining the ordering by making permutations of the candidate terms-and selecting a permutated ordering that has a highest count in a database containing reliable DEN entries.

13. The method of claim 1, wherein creating the candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM includes determining the ordering by making permutations of the candidate terms and selecting a permutated ordering that has a highest count in search engine results.

14. The method of claim 1, further comprising, validating the candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM by comparison of the ordering with reliable DEN entries in a database.

15. The method of claim 1, further comprising, validating the candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM by usage counts in search engine results,

16. A system, comprising:

a database of names of data elements used in data structures, formats and models; and

a dictionary entry name (DEN) generator hosted on a computer and coupled to the database,

wherein the DEN generator is configured to analyze the names of data elements to identify candidate terms for a DEN of a node or equivalence class of data elements in a canonical data model (CDM).

17. The system of claim 16, wherein DEN generator is configured to receive input data on terms used in data structures, formats and models, and identify candidate terms for the DEN of the node or equivalence class in the CDM by linguistics and semantic analysis of terms in the input data.

18. The system of claim 17, wherein DEN generator is configured to revise or validate an ordering of the candidate terms for the DEN of the node or equivalence class in the CDM by comparing the ordering with known DENs stored in a database.

19. A computer program product embodied in non-transitory computer-readable media that includes executable code, which when, executed:

identifies candidate terms for a dictionary entry name (DEN) of a node or equivalence class of data elements in a canonical data model;

counts a frequency of occurrence of candidate terms in a data store; and

creates a candidate ordering of terms for the complete ordered DEN of the node or

equivalence class of the CDM based at least in part on the counts of frequency of occurrence.

20. The computer program product of claim 19, which further includes executable code, which when executed:

validates the candidate ordering of terms for the complete ordered DEN of the node or equivalence class of the CDM by comparison of the ordering with reliable DEN entries in a database and/or by usage counts in search engine results.