Multi-lingual recognizing method using context information
US5995919A · kind A · utility
Assignee
Inventors
Key dates
| Filing date | Jul 24, 1997 |
| Grant date | Nov 30, 1999 |
| Priority date | — |
| Expiry date | Jul 24, 2017 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06F40/40
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
An automatic multi-lingual recognizing method for distinguishing multi-lingual characters among different codes is disclosed. The method includes determining the code legitimacy of characters of a whole text by comparing code blocks of the characters with predefined code ranges of code systems. Next, if the code legitimacy of the characters is not uniquely recognized, the characters of the whole text is analyzed to determine whether the characters are common phrases by comparing the characters with common-phrase data bases. Finally, if the analysis of the common phrase of the characters is not uniquely recognized, the characters of the text is then analyzed to determine whether the characters are common words by comparing the characters with common-word data bases.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.