| Summary: | 碩士 === 逢甲大學 === 資訊工程所 === 99 === Over the past 10 years, the most popular leisure activity in the world is to play online games. There are more and more people in the online game market due to it being mature in industry and rapid growth market. Meanwhile, there are more and more people who tend to use game bot to control the games instead of doing it manually. This is the most serious threat which will affect the online game market, because game bot users can obtain a certain amount of gains, but they do not need to put forth any effort. It is obviously unfair to other players who are playing the games without game bot. If game bot is being rampant, it will lead to economical collapse in the online game industry and it will shorten its life cycle. Most of the gaming companies would detect game bot users manually. However, it would be a waste of labor. Therefore, instead of manpower the gaming corporations would like to use software to detect those users in order to fight against game bot and to maintain a high-quality Online Game environment. In this paper, we are going to collect the gaming trace and use Dynamic Bayesian Network to examine the cheat probability curve of graphs using game bot as the player. If the curves cross the threshold, it will be determined that the player is game bot. We use online rhythm game as the experiment environment. Through the proposed scheme we discussed as above, we can filtered a large number of suspect players.
|
|---|
