Problem to be solved: to provide a multilingual speech recognition method in which accuracy to transform into sentence is improved.The method isPerformed by the speech recognition sectionGet the alphabet string corresponding to the speech input signal according to the alphabet tableThe alphabet string is determined to correspond to multiple languages according to the multilingual vocabularyForming a sentence according to multilingual vocabulary and original wordsExecuted by semantic analysisSentence and theme Vocabulary - according to the semantic data setA modified procedure to generate a modified sentenceAn analysis state determination procedure or a procedure for outputting a sentence is selectively executed;When the fix procedure succeedsOutput a corrected sentenceIf the modification procedure failedPerform an analysis state determination procedureThe analysis state determining procedure for selectively outputting the determined result is executed.Diagram
展开▼