Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning

—Game testing has been long recognized as a notoriously challenging task, which mainly relies on manual playing and scripting based testing in game industry. Even until recently, automated game testing still remains to be largely untouched niche. A key challenge is that game testing often requires t...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHENG, Yan, XIE, Xiaofei, SU, Ting, MA, Lei, HAO, Jianye, MENG, Zhaopeng, LIU, Yang, SHEN, Ruimin, CHEN, Yingfeng, FAN, Changjie
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2019
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/7065
https://ink.library.smu.edu.sg/context/sis_research/article/8068/viewcontent/e38c5f3d60a830785f1bdd8b69563c45.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-8068
record_format dspace
spelling sg-smu-ink.sis_research-80682022-04-07T08:18:50Z Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning ZHENG, Yan XIE, Xiaofei SU, Ting MA, Lei HAO, Jianye MENG, Zhaopeng LIU, Yang SHEN, Ruimin CHEN, Yingfeng FAN, Changjie —Game testing has been long recognized as a notoriously challenging task, which mainly relies on manual playing and scripting based testing in game industry. Even until recently, automated game testing still remains to be largely untouched niche. A key challenge is that game testing often requires to play the game as a sequential decision process. A bug may only be triggered until completing certain difficult intermediate tasks, which requires a certain level of intelligence. The recent success of deep reinforcement learning (DRL) sheds light on advancing automated game testing, without human competitive intelligent support. However, the existing DRLs mostly focus on winning the game rather than game testing. To bridge the gap, in this paper, we first perform an in-depth analysis of 1349 real bugs from four real-world commercial game products. Based on this, we propose four oracles to support automated game testing, and further propose Wuji, an on-the-fly game testing framework, which leverages evolutionary algorithms, DRL and multi-objective optimization to perform automatic game testing. Wuji balances between winning the game and exploring the space of the game. Winning the game allows the agent to make progress in the game, while space exploration increases the possibility of discovering bugs. We conduct a large-scale evaluation on a simple game and two popular commercial games. The results demonstrate the effectiveness of Wuji in exploring space and detecting bugs. Moreover, Wuji found 3 previously unknown bugs1 , which have been confirmed by the developers, in the commercial games 2019-11-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/7065 info:doi/10.1109/ASE.2019.00077 https://ink.library.smu.edu.sg/context/sis_research/article/8068/viewcontent/e38c5f3d60a830785f1bdd8b69563c45.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Testing Artificial Intelligence Deep Reinforcement Learning Evolutionary Multi-Objective Optimizatio Software Engineering
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic Testing
Artificial Intelligence
Deep Reinforcement Learning
Evolutionary Multi-Objective Optimizatio
Software Engineering
spellingShingle Testing
Artificial Intelligence
Deep Reinforcement Learning
Evolutionary Multi-Objective Optimizatio
Software Engineering
ZHENG, Yan
XIE, Xiaofei
SU, Ting
MA, Lei
HAO, Jianye
MENG, Zhaopeng
LIU, Yang
SHEN, Ruimin
CHEN, Yingfeng
FAN, Changjie
Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning
description —Game testing has been long recognized as a notoriously challenging task, which mainly relies on manual playing and scripting based testing in game industry. Even until recently, automated game testing still remains to be largely untouched niche. A key challenge is that game testing often requires to play the game as a sequential decision process. A bug may only be triggered until completing certain difficult intermediate tasks, which requires a certain level of intelligence. The recent success of deep reinforcement learning (DRL) sheds light on advancing automated game testing, without human competitive intelligent support. However, the existing DRLs mostly focus on winning the game rather than game testing. To bridge the gap, in this paper, we first perform an in-depth analysis of 1349 real bugs from four real-world commercial game products. Based on this, we propose four oracles to support automated game testing, and further propose Wuji, an on-the-fly game testing framework, which leverages evolutionary algorithms, DRL and multi-objective optimization to perform automatic game testing. Wuji balances between winning the game and exploring the space of the game. Winning the game allows the agent to make progress in the game, while space exploration increases the possibility of discovering bugs. We conduct a large-scale evaluation on a simple game and two popular commercial games. The results demonstrate the effectiveness of Wuji in exploring space and detecting bugs. Moreover, Wuji found 3 previously unknown bugs1 , which have been confirmed by the developers, in the commercial games
format text
author ZHENG, Yan
XIE, Xiaofei
SU, Ting
MA, Lei
HAO, Jianye
MENG, Zhaopeng
LIU, Yang
SHEN, Ruimin
CHEN, Yingfeng
FAN, Changjie
author_facet ZHENG, Yan
XIE, Xiaofei
SU, Ting
MA, Lei
HAO, Jianye
MENG, Zhaopeng
LIU, Yang
SHEN, Ruimin
CHEN, Yingfeng
FAN, Changjie
author_sort ZHENG, Yan
title Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning
title_short Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning
title_full Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning
title_fullStr Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning
title_full_unstemmed Wuji: Automatic online combat game testing using evolutionary deep reinforcement learning
title_sort wuji: automatic online combat game testing using evolutionary deep reinforcement learning
publisher Institutional Knowledge at Singapore Management University
publishDate 2019
url https://ink.library.smu.edu.sg/sis_research/7065
https://ink.library.smu.edu.sg/context/sis_research/article/8068/viewcontent/e38c5f3d60a830785f1bdd8b69563c45.pdf
_version_ 1770576197896372224