Game Bot Identification Based on Manifold Learning
2 likes1,321 views
This document summarizes a research paper on identifying game bots based on manifold learning. It discusses how game bots can negatively impact online games and how they are currently detected through human analysis of behavior patterns. The paper then proposes a new method that constructs pace vectors from avatar movement trajectories and applies manifold learning and classification to distinguish human from bot trajectories. It tests this approach on Quake 2 game data and achieves high accuracy even when noise is added to bot trajectories. The method is also shown to generalize across different maps.
![Related Work
Prevention
CAPTCHA
(reverse Turing
tests) [Golle et al; 2005]
Detection
Process monitoring at client side [GameGuard]
Bot program’s signatures are keeping changing
Traffic analysis at the network [Chen et al; 2006]
Remove bot traffic’s regularity by heavy‐tailed random delays
Aiming bot detection using DBN [Yeung et al; 2006]
Specific to aiming bots that help aim the target accurately
Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 4](https://image.slidesharecdn.com/chen08manifoldslides-100208024203-phpapp02/85/Game-Bot-Identification-Based-on-Manifold-Learning-4-320.jpg)
Game Bot Identification Based on Manifold Learning
- 1. Game Bot Identification based on Manifold Learning Kuan‐Ta Chen Academia Sinica Hsing‐Kuo Pao NTUST Hong‐Chung Chang NTUST ACM NetGames 2008
- 2. Game Bots Game bots: automated AI programs that can perform certain tasks in place of gamers Popular in MMORPG and FPS games MMORPGs (Role Playing Games) accumulate rewards in 24 hours a day break the balance of power and economies in game FPS games (First‐Person Shooting Games) a) improve aiming accuracy only b) fully automated achieve high ranking without proficient skills and efforts Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 2
- 3. Bot Detection Detecting whether a character is controlled by a bot is difficult since a bot obeys the game rules perfectly No general detection methods are available today State of practice is identifying via human intelligence Detect by “bots may show regular patterns or peculiar behavior” Confirm by “bots cannot talk like humans” Labor‐intensive and may annoy innocent players Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 3
- 4. Related Work Prevention CAPTCHA (reverse Turing tests) [Golle et al; 2005] Detection Process monitoring at client side [GameGuard] Bot program’s signatures are keeping changing Traffic analysis at the network [Chen et al; 2006] Remove bot traffic’s regularity by heavy‐tailed random delays Aiming bot detection using DBN [Yeung et al; 2006] Specific to aiming bots that help aim the target accurately Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 4
- 5. CAPTCHA in a Japanese Online Game Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 5
- 6. Our Goal of Bot Detection Solutions Passive detection No intrusion in players’ gaming experience No client software support is required Generalizable schemes (for other games and other game genres) Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 6
- 7. Our Solution: Trajectory + Manifold Learning Based on the avatar’s movement trajectory in game Applicable for all genres of games where players control the avatar’s movement directly Avatar’s trajectory is high‐dimensional (both in time and spatial domain) Use manifold learning to distinguish the trajectories of human players and game bots Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 7
- 8. The Rationale behind Our Scheme The trajectory of the avatar controlled by a human player is hard to simulate for two reasons: Complex context information: Players control the movement of avatars based on their knowledge, experience, intuition, and a great deal of environmental information in game. Human behavior is not always logical and optimal How to model and simulate realistic movements (for game agents) is still an open question in the AI field. Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 8
- 9. Bot Detection: A Decision Problem Q: Whether a bot is controlling a game client given the movement trajectory of the avatar? A: Yes / No? Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 9
- 10. Talk Progress Overview Data Description Proposed Scheme Pace vector construction Dimension Reduction using Isomap Classification Performance Evaluation Conclusion Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 10
- 11. Case Study: Quake 2 Choose Quake 2 as our case study A classic FPS game Many real‐life human traces are available on the Internet more realistic than traces collected in experiments Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 11
- 13. Data Collection Human traces downloaded from fan sites including GotFrag Quake, Planet Quake, Demo Squad, and Revilla Quake Site Bot traces collected on our own Quake server CR BOT 1.14 Eraser Bot 1.01 ICE Bot 1.0 Totally 143.8 hours of traces were collected Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 13
- 14. Aggregate View of Trails (Human & 3 Bots) Human CR Bot Eraser ICE Bot
- 18. Talk Progress Overview Data Description Proposed Scheme Pace vector construction Dimension Reduction using Isomap Classification Performance Evaluation Conclusion Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 18
- 19. The Complete Process: Overview Step 1. Pace Vector Construction Step 2. Dimension Reduction with Isomap Step 3. Supervised classification Decision Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 19
- 20. Step 1. Pace Vector Construction For each trace sn , we compute the pace (distance) in successive two seconds by p ksn,i+1 − sn,i k = (sn,i+1 − sn,i )T (sn,i+1 − sn,i ) We then compute the distribution (histogram) of paces with a fixed bin size by Fn = (fn,1 , fn,2 , . . . , fn,B ) where B is the number of bins in the distribution. Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 20
- 21. Pace Vector: An Example B is set to 200 (dimensions) in this work Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 21
- 22. Step 2. Dimension Reduction with Isomap We adopt Isomap for nonlinear dimension reduction for Better classifiaction accuracy Lower computation overhead in classification Isomap Assume data points lie on a manifold A mathematical space in which every point has a neighborhood which resembles Euclidean space, but in which the global structure may be more complicated. (Wikipedia) 1. Construct the neighborhood graph by kNN (k‐nearest neighbor) 2. Compute the shortest geodesic path for each pair of points 3. Reconstruct data by MDS (multidimensional scaling) Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 22
- 23. A Graphic Representation of Isomap Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 23
- 25. Step 3. Classification Apply a supervised classifier on the Isomap‐reduced pace vectors SVM (Support Vector Machine) in our study To decide whether a trajectory belongs to a game bot or a human player Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 25
- 26. Talk Progress Overview Data Description Proposed Scheme Pace vector construction Dimension Reduction using Isomap Classification Performance Evaluation Conclusion Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 26
- 27. Five Methods for Comparison Method Data Input kNN Original 200‐dimension Linear SVM Pace Vectors Nonlinear SVM Isomap + kNN Isomap‐reduced Pace Isomap + Nonlinear SVM Vectors Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 27
- 28. Evaluation Results Error Rate False Postive Rate False Negative Rate
- 29. Addition of Gaussian Noise Bot programmers can try to evade from detection by adding random noise into bots’ movement behavior Evaluate the robustness of our schemem by adding Gaussian noise into bots’ trajectories Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 29
- 30. Evaluation Results Error Rate False Postive Rate False Negative Rate Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 30
- 31. Cross‐Map Validation Human movement may be restricted by the environment around him/her Whether a classifier trained for a map can be used for detecting bots on another map? The Edge The Frag Pipe Warehouse Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 31
- 32. Evaluation Results Error Rate False Postive Rate False Negative Rate
- 33. Talk Progress Overview Data Description Proposed Scheme Pace vector construction Dimension Reduction using Isomap Classification Performance Evaluation Conclusion Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 33
- 34. Conclusion We propose a trajectory‐based approach for detecting game bots. The results show that the Isomap + nonlinear SVM approach performs good and stable results. Human’s logic in controlling avatars is hard to simulate we believe this approach has the potential to be a general yet robust bot detection methodology Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 34
- 35. Future Work Include more spatial‐domain information in the pace vector Validate our methodology on other games (game genres) Kuan‐Ta Chen / Game Bot Identification based on Manifold Learning 35
- 36. Thank You! Kuan‐Ta Chen http://www.iis.sinica.edu.tw/~ktchen ACM NetGames 2008