| 摘要 |
PURPOSE:To improve the use efficiency of a personal dictionary by replacing gradually a common dictionary with the personal dictionary with regard to a dictionary, as the number of characters recorded by a specified writer increases. CONSTITUTION:A dictionary 2 stores the statistic value at every category related to a feature pattern extracted from a character pattern to which the specified user of this device records. For instance, when feature patterns f1, f2,...fn1 whose character category is 'sea', feature patterns g1, g2,...gn2 whose character category is 'mountain', and feature patterns h1, h2,...hn3 whose character category is 'river' are inputted to a learning part 7, the learning part 7 outputs average values (f), (g), and (h) as standard patterns related to 'sea', 'mountain', and 'river'. A dictionary replacing part 8 decides whether each standard pattern belonging to the dictionary 2 can be used enough for a dictionary 1 for recognition, and when it is decided to be sufficient, the dictionary is replaced by a standard pattern unit.
|
