Analyzing multilingual diagnostic records

• US 10,062,222 B2
• Filed: 11/23/2015
• Issued: 08/28/2018
• Est. Priority Date: 11/23/2015
• Status: Active Grant

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: GM Global Technology Operations LLC (General Motors Company)
• Inventors: Rajpathak, Dnyanesh G., De, Soumen
• Primary Examiner(s): Roberts, Shaun
• Application Number:
  US14/949,614
• Publication Number:
  US 20170148233A1
• Time in Patent Office: 1,009 Days
• Field of Search: 704 9, 704 10, 701 291
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G07C 5/00 Registering or indicating the working of vehicles
• CPC Class Codes: G07C 5/006 Indicating maintenance
• CPC Class Codes: G07C 5/0808 Diagnosing performance data testing of vehicles G01M17/00; testing of electrical installation on vehicles G01R31/005

Abstract

A system and method of analyzing content of multilingual vehicle diagnostic records includes: determining a word window within a vehicle diagnostic record; identifying a pair or a tuple comprising parts, symptoms, or actions; generating a plurality of pairs or tuples comprising parts, symptoms, or actions; determining a frequency value for each pair or tuple; and comparing the determined frequency value with a predetermined threshold.

Claims

1. A method of analyzing content of multilingual vehicle diagnostic records to identify data features in the vehicle diagnostic records independent of the language of the diagnostic record, the method comprising the steps of:

(a) electronically receiving at a central facility computer multilingual vehicle diagnostic records;

(b) using a trained database, classifying one or more words within each of the vehicle diagnostic records as a part, a symptom, or an action based on the position of each word within the vehicle diagnostic records;

(c) determining a word window within each vehicle diagnostic record that identifies particular parts, symptoms, or actions;

(d) using the trained database, determining relationships between the parts, symptoms, or actions based on pairs or tuples formed from combinations of the parts, symptoms, or actions within the word window;

(e) determining a frequency value for each pair or tuple using the trained database;

(f) comparing the determined frequency value with a predetermined threshold, and identifying each pair or tuple as a data feature of the diagnostic record when the determined frequency value is equal to or above the predetermined threshold; and

(g) storing the data features identified in step (f) for each of the multilingual vehicle diagnostic records received in step (a), wherein the data features enable analysis of content in the vehicle diagnostic records independent of the content language.

2. The method of claim 1, further comprising the step of identifying a focal word within the word window.

3. The method of claim 1, further comprising the step of deleting one or more pairs or tuples when the determined frequency value is not equal to or above the predetermined threshold.

4. The method of claim 1, wherein the frequency value is obtained by determining the number of times the pairs or tuples appear in the trained database.

5. The method of claim 1, wherein step (c) further comprises forming the tuples based on content from the pairs.

15. The method of claim 1, further comprising clustering vehicle diagnostic data using the data features.

16. The method of claim 15, wherein clustering vehicle diagnostic data using the data features includes clustering combinations of data features relating to parts, symptoms, failure modes, or actions.

6. A method of analyzing content of multilingual vehicle diagnostic records to identify data features in the vehicle diagnostic records independent of the language of the diagnostic record, the method comprising the steps of:

(a) electronically receiving at a central facility computer multilingual vehicle diagnostic records;

(b) identifying a position for each of the words in the vehicle diagnostic records;

(c) using a trained database, classifying one or more words within each of the vehicle diagnostic records as a part, a symptom, or an action based on the position of each word;

(d) identifying a focal word in each vehicle diagnostic record;

(e) determining a word window that includes the focal word;

(f) generating a plurality of unique pairs or tuples representing combinations of parts, symptoms, or actions within the word window;

(g) determining a frequency value for each unique pair or tuple using a trained database;

(h) comparing the determined frequency value with a predetermined threshold, and identifying each unique pair or tuple as a data feature of the diagnostic record when the determined frequency value is equal to or above the predetermined threshold; and

(i) storing the data features identified in step (h) for each of the multilingual vehicle diagnostic records received in step (a), wherein the data features enable analysis of content in the vehicle diagnostic records independent of language.

7. The method of claim 6, further comprising the step of deleting one or more pairs or tuples when the determined frequency value is not equal to or above the predetermined threshold.

8. The method of claim 6, wherein the pair or tuple comprising part, symptoms, or actions includes words found within the word window.

9. The method of claim 6, wherein the frequency value is obtained by determining the number of times the pairs or tuples appear in the trained database.

10. The method of claim 6, wherein step (d) further comprises generating the tuples based on content from the pairs.

11. A method of analyzing content of multilingual vehicle diagnostic records to identify data features in the vehicle diagnostic records independent of the language of the diagnostic record, the method comprising the steps of:

(a) electronically receiving at a central facility computer multilingual vehicle diagnostic records;

(b) obtaining classifications from a database trained to classify words according to the position of the words in the vehicle diagnostic records;

(c) classifying one or more words within each of the vehicle diagnostic records as a part, a symptom, or an action based on the position of each of the one or more words within the vehicle diagnostic records by applying the classifications obtained in step (b);

(d) identifying a focal word in each of the vehicle diagnostic records;

(e) determining a word window that includes the focal word;

(f) generating a plurality of unique pairs or tuples representing combinations of parts, symptoms, or actions within the word window;

(g) determining a frequency value for each unique pair or tuple using a trained database;

(h) comparing the determined frequency value with a predetermined threshold, and identifying each unique pair or tuple as a data feature of the diagnostic record when the determined frequency value is equal to or above the predetermined threshold; and

(i) storing the data features identified in step (h) for each of the multilingual vehicle diagnostic records received in step (a), wherein the data features enable analysis of content in the vehicle diagnostic records independent of language.

12. The method of claim 11, further comprising the step of deleting one or more pairs or tuples when the determined frequency value is not equal to or above the predetermined threshold.

13. The method of claim 11, wherein the frequency value is obtained by determining the number of times the pairs or tuples appear in the trained database.

14. The method of claim 11, wherein step (f) further comprises generating the tuples based on content from the pairs.