RECOGNIZING CHEMICAL NAMES IN A CHINESE DOCUMENT

• US 20130054226A1
• Filed: 08/30/2012
• Published: 02/28/2013
• Est. Priority Date: 08/31/2011
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Chen, Ying, Su, Zhong, Wu, Xian, Zhang, Li
• Granted Patent: US 9,575,957 B2
• Time in Patent Office: Days
• US Class Current: 704/8
• CPC Class Codes: G06F 40/279 Recognition of textual entities

Abstract

A method and system for recognizing chemical names in a Chinese document. The method includes: receiving a Chinese document including chemical names; recognizing chemical name segments in the document; recognizing non-chemical name segments in the document; and combining the chemical name segments to get chemical names based on the recognized chemical name segments and non-chemical name segments. Specific embodiments of the present invention can effectively recognize chemical names from a chemical document.

Claims

1. A method for recognizing chemical names in Chinese documents, comprising:

receiving a Chinese document including chemical names;

recognizing chemical name segments in said document;

recognizing non-chemical name segments in said document; and

combining said chemical name segments to get said chemical names based on said recognized chemical name segments and non-chemical name segments.

2. The method according to claim 1, wherein said non-chemical name segments are used as separators in combining said chemical name segments to get said chemical names.

3. The method according to claim 1, wherein recognizing said chemical name segments in said document comprises:

segmenting said document into sentences;

matching all of said chemical name segments appearing in sentences of said document based on a chemical name segment dictionary;

recording position information of said chemical name segments; and

reducing said chemical name segments in a same sentence.

4. The method according to claim 3, wherein recognizing said non-chemical name segments in said document comprises:

segmenting said document into words;

checking whether each segmented word is in a non-chemical name segment dictionary;

provided that said segmented word is in said non-chemical name segment dictionary, determining said segmented word to be a non-chemical name segment; and

recording position information of said non-chemical name segment.

5. The method according to claim 4, wherein combining said chemical name segments to get said chemical name based on said recognized chemical name segments and non-chemical name segments comprises:

determining adjacent chemical name segments in a same sentence according to said position information of said chemical name segments;

checking whether there are non-chemical name segments between said adjacent chemical name segments based on said position information of said chemical name segments and non-chemical name segments; and

provided that there are no non-chemical name segments between said adjacent chemical name segments, combining said adjacent chemical name segments to get a chemical name.

6. The method according to claim 1, further comprising:

extending towards both sides from positions of said chemical name segments until meeting non-chemical name segments, so as to obtain the chemical names.

7. The method according to claim 4, wherein:

said position information includes the start positions and end positions of one of said chemical name segments and non-chemical name segments in said document.

8. The method according to claim 4, further comprising at least one:

generating said chemical name segment dictionary from existing Chinese chemical names; and

forming said non-chemical name segment dictionary based on words with high frequency collected from Chinese documents in non-chemical fields.

9. The method according to claim 3, wherein reducing said chemical name segments in a same sentence is performed according to a principle of matching the most chemical name segments with the least number of chemical name segments.

10. The method according to claim 8, further comprising:

generating said chemical name segment dictionary from existing English chemical names.

11. A system for recognizing chemical names in a Chinese document, comprising:

receiving means configured to receive a Chinese document including chemical names;

first recognizing means configured to recognize chemical name segments in said document;

second recognizing means configured to recognize non-chemical name segments in said document; and

combining means configured to combine said chemical name segments to get said chemical names based on said recognized chemical name segments and non-chemical name segments.

12. The system according to claim 11, wherein said non-chemical name segments are used as separators in combining said chemical name segments to get said chemical names

13. The system according to claim 11, wherein said first recognizing means comprises:

sentence-segmenting means configured to segment said document into sentences;

matching means configured to match all of said chemical name segments appearing in sentences of said document based on a chemical name segment dictionary;

first recording means configured to record position information of said chemical name segments; and

reducing means configured to reduce said chemical name segments in a same sentence.

14. The system according to claim 13, wherein said second recognizing means comprises:

word-segmenting means configured to segment said document into words;

first checking means configured to check whether each segmented word is in a non-chemical name segment dictionary;

first determining means configured to, provided that said segmented word is in said non-chemical name segment dictionary, determine said segmented word to be a non-chemical name segment; and

second recording means configured to record position information of said non-chemical name segment.

15. The system according to claim 14, wherein said combining means comprises:

second determining means configured to determine adjacent chemical name segments in a same sentence according to said position information of said chemical name segments;

second checking means configured to check whether there are non-chemical name segments between said adjacent chemical name segments based on said position information of said chemical name segments and non-chemical name segments; and

combination executing means configured to, provided that there are no non-chemical name segments between said adjacent chemical name segments, combine said adjacent chemical name segments to get a chemical name.

16. The system according to claim 11, wherein said system further comprises:

extending means configured to extend towards both sides from the position of said chemical name segments until meeting non-chemical name segments, so as to obtain the chemical names.

17. The system according to 14, wherein:

said position information includes the start positions and end positions of one of said chemical name segments and non-chemical name segments in said document.

18. The system according to claim 14, further comprising at least one of:

chemical name segment dictionary generating means configured to generate said chemical name segment dictionary from existing Chinese chemical names;

or

non-chemical name segment dictionary generating means configured to form said non-chemical name segments dictionary based on words with high frequency collected from Chinese documents in non-chemical fields.

19. The system according to claim 13, wherein reducing said chemical name segments in a same sentence is performed according to a principle of matching the most chemical name segments with the least number of chemical name segments.

20. The system according to claim 18, wherein said chemical name segment dictionary generating means generates said chemical name segment dictionary from existing English chemical names.