| 摘要 |
PURPOSE: A method for recognizing a voice by considering a phrase structure and a recording medium thereof are provided to improve the learning efficiency of speaking through repetitive learning. CONSTITUTION: A voice recognition device determines a first meaning object by recognizing a voice inputted by a user(S100). The device receives the first meaning object to search for a second meaning object candidate sequentially connected with the first meaning object(S200). If a connection candidate is determined for the meaning objects, the device compares the meaning objects with each other(S400). [Reference numerals] (AA) Start; (BB) End; (S100) Determining a first meaning object; (S200) Determining a second meaning object candidate; (S300) Are all meaning objects processed?; (S400) Comparing the meaning objects
|
