Normalization of medical terms with multi-lingual resources

Patent metadata

Field	Value
Publication number	US-11308289-B2
Application number	US-201916569874-A
Country	US
Kind code	B2
Filing date	Sep 13, 2019
Priority date	Sep 13, 2019
Publication date	Apr 19, 2022
Grant date	Apr 19, 2022

Abstract

Method and apparatus are presented for receiving a medical or medical condition related input term or phrase in a source language, and translating the term or phrase from the source language into at least one target language to obtain a set of translated terms of the input term. For each translated term in the set of translations, the method and apparatus further translate the set of translations back into the source language to obtain an output list of standard versions of the input term, scoring each entry of the output list as to probability of being the most standard version of the input term, and providing the entry of the output list that has the highest score to a user.

First claim

What is claimed is:

1. A method comprising:
receiving data including a medically-related input term in a source language;
translating, using a first dictionary, the input term from the source language into a plurality of target languages to obtain a set of translated terms of the input term;
for each respective translated term in the set of translated terms, translating, using a second dictionary, from one or more pre-defined domain specific dictionaries, that is specific to a medical domain, the respective translated term back into the source language to obtain an output list of normalized versions of the input term, wherein each normalized version in the output list are specific to the medical domain;
assigning a respective score to each respective entry of the output list of normalized versions based on a respective probability of the respective entry being the most standard version of the input term;
identifying an entry of the output list having a highest score; and
providing the entry of the output list having the highest score as normalized input to a medical data processing application that processes the data.

2. The method of claim 1 , wherein the input term is a common version of a medical concept, condition, object, process, or drug.

3. The method of claim 1 , wherein each element of the set of translations is translated back into the source language using the one or more pre-defined domain specific dictionaries.

4. The method of claim 3 , wherein the one or more pre-defined domain specific dictionaries are professional, scientific, or academic.

5. The method of claim 1 , wherein the input term is used in spoken Chinese.

6. The method of claim 1 , wherein the plurality of target languages includes two separate target languages.

7. The method of claim 6 , wherein the two separate target languages are each European languages.

8. The method of claim 6 , wherein, for each of the two separate target languages a body of published medical literature exists.

9. The method of claim 1 , further comprising outputting each entry of the output list, and its score, to the medical data processing application.

10. The method of claim 1 , wherein the score of each entry of the output list is determined by one or more of:
relative frequency of the entry in the source language, relative frequency of a term in the plurality of target languages that was translated into the entry, or relative authority or prestige of the dictionary that was used to obtain the entry.

11. A computer program product comprising a computer-readable storage medium having computer-readable program code embodied therewith, the computer-readable program code executable by one or more computer processors to perform an operation comprising:
receiving data including a medically-related input term in a source language;
translating, using a first dictionary, the input term from the source language into a plurality of target languages to obtain a set of translated terms of the input term;
for each respective translated term in the set of translated terms, translating, using a second dictionary, from one or more pre-defined domain specific dictionaries, that is specific to a medical domain, the respective translated term back into the source language to obtain an output list of normalized versions of the input term, wherein each normalized version in the output list are specific to the medical domain;
assigning a respective score to each respective entry of the output list of normalized versions based on a respective probability of the respective entry being the most standard version of the input term;
identifying an entry of the output list having a highest score; and
provide the entry of the output list having the highest score as normalized input to a medical data processing application that processes the data.

12. The computer program product of claim 11 , wherein the computer-readable program code is further executable to:
output each entry of the output list, and its score, to the medical data processing application.

13. The computer program product of claim 11 , wherein each element of the set of translations is translated back into the source language using the one or more pre-defined domain specific dictionaries.

14. The computer program product of claim 13 , wherein the one or more pre-defined domain specific dictionaries are professional, scientific, or academic.

15. The computer program product of claim 11 , wherein the score of each entry of the output list is determined by one or more of:
relative frequency of the entry in the source language, relative frequency of a term in the plurality of target languages that was translated into the entry, or relative authority or prestige of the dictionary that was used to obtain the entry.

16. A system, comprising:
one or more computer processors; and
a memory containing a program which when executed by the one or more computer processors performs an operation, the operation comprising:
receiving data including a medically-related input term in a source language;
translating, using a first dictionary, the input term from the source language into a plurality of target languages to obtain a set of translated terms of the input term;
for each respective translated term in the set of translated terms, translating, using a second dictionary, from one or more pre-defined domain specific dictionaries, that is specific to a medical domain, the respective translated term back into the source language to obtain an output list of normalized versions of the input term, wherein each normalized version in the output list are specific to the medical domain;
assigning a respective score to each respective entry of the output list of normalized versions based on a respective probability of the respective entry being the most standard version of the input term;
identifying an entry of the output list having a highest score; and
providing the entry of the output list having the highest score as normalized input to a medical data processing application that processes the data.

17. The system of claim 16 , the operation further comprising:
outputting each entry of the output list, and its score, to the medical data processing application.

18. The system of claim 16 , wherein each element of the set of translations is translated back into the source language using the one or more pre-defined domain specific dictionaries.

19. The system of claim 18 , wherein the one or more pre-defined domain specific dictionaries are professional, scientific, or academic.

20. The system of claim 16 , wherein the score of each entry of the output list is determined by one or more of:
relative frequency of the entry in the source language, relative frequency of a term in the plurality of target languages that was translated into the entry, or relative authority or prestige of the dictionary that was used to obtain the entry.

Assignees

• IBM

Inventors

• Jiang Jian Min
• Wang Jian
• Huang Songfang
• Li Jing
• Wang Ke

Classifications

• G06F40/58 Primary

  Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation · CPC title

• G06F40/197

  Version control (for software G06F8/71) · CPC title

• G06F40/51

  Translation evaluation · CPC title

• G06F40/44

  Statistical methods, e.g. probability models · CPC title

• G06F40/237

  Lexical tools · CPC title