| 摘要 |
<p>A method for forming a game character is provided to enable an intelligent character to learn reaction to an action pattern of an opponent character quickly in a match action game without making a game developer perform coding. A pool of intelligent characters used by a game user is prepared(S100). The pool of opponent characters having a match with the intelligent character is prepared(S110). The intelligent character is upgraded by making the intelligent character have the match with all opponent characters in a game(S120). The intelligent character is upgraded by learning a match result with the opponent characters. Learning is performed by storing a feature of the intelligent character when the intelligent character obtains a higher score than the opponent character.</p> |
