SlideShare a Scribd company logo
1 of 16
Intelligent Adaptation of Difficulty and NPC Behavior
in Serious Video Games for Learning
Ivan Naydenov, Ilko Adamov, Boyan Bontchev
SUMMIT Annual Conference
Sofia, Bulgaria
April 24, 2024
Contract BG-RRP-2.004-0008 for the financing of project "Sofia University - Marking Momentum for Innovation and Technological Transfer“ under pillar 2 "Establishing a network of research higher education institutions in Bulgaria",
component "Innovative Bulgaria" from National Recovery and Resilience Plan part of the program to accelerate economic recovery and transformation through science and innovation
Two key issues of user-centric adaptation of
video games for learning
• Tailoring the learning task difficulty – it is very important because didactic
tasks are mapped to gaming tasks;
• Adapting the behavior of non-player characters (NPCs), i.e., virtual heroes –
applied in educational games for:
• (a) teaching assistants (the most popular case);
• (b) concurrent learners; or
• (c) opposite characters or enemies.
Recognizing emotions from facial expressions
• Understanding human behavior: recognizing emotions from facial expressions allows us
to gain insight into individual and collective emotional states.
• Communication and social interaction: facial expressions are a fundamental aspect of
nonverbal communication.
• User experience and human-computer interaction: recognizing emotions from facial
expressions is particularly relevant in the field of human-computer interaction – systems
and interfaces can be designed to adapt and respond accordingly.
• Psychological and mental health assessment: the ability to recognize and measure
emotions from facial expressions can be valuable in psychological and mental health
assessment.
Popular methods for emotion analysis
A. By using facial expressions:
• Classification using machine learning models like KNN, SVM, etc.
• Convolutional Neural Networks (CNNs)
B. Measurement and analysis of physiological data like Electrocardiograph
(ECG), Blood Volume Pulse (BVP), Galvanic Skin Response (GSR) and
Electromyography (EMG)
• Measure and extract physiological data
• Analyze and recognize emotions by data clustering
Recognition of player emotional state by CNN
• Colab (https://colab.research.google.com) with various Python libraries for data
analysis and visualization
• The neural network architecture consists of 779,718 neurons organized into 17
layers divided into four blocks.
• Each of the first three blocks contains three convolutional layers and one pooling
layer.
• Convolution layers employ filters or kernels to extract local features, such as edges,
corners, and textures, from the input image.
• Pooling layers are applied after convolution to reduce the spatial dimensions of the
feature maps using techniques like max pooling or average pooling.
• The final block includes a fully connected layer, which combines the high-level
features extracted from previous layers to produce the final classification output.
Training and validation accuracy (left) and loss
(right) of the neural network
Confusion matrix of
neural network
Samples of BVP (left) and GSR (right) data – used for data clusterization
K-Means (left), mean-shift (middle), and
agglomerative (right) clustering based on
physiological data
Adaptation of difficulty and NPC tutor behavior
The player:
• In situation 1 – has played a lot without any engagement and motivation;
• In situation 2 – has achieved high score and demonstrated good skills but remains in apathy;
• In situation 3 – is disappointed by his/her low outcome;
• In situation 4 – manages to play well but at the price of some non-desired emotions;
• In situation 5 – is happy to play the game but without trying to achieve good outcomes;
• In situation 6 – is happy to play the educational game while achieving a good score;
• In situation 7 – cannot achieve a good score and has both desired and non-desired emotions;
• In situation 8 – has succeeded in the game but at the price of negative emotions.
Situa-
tion
Desired
Emotion
Joy, Anger
Non-desired
emotion
Fear, Sadness
Out-
comes Difficulty
NPC
tutor
behavior
1 Low Low Low Const Encouraging
2 Low Low High Increase Satisfied, encouraging
3 Low High Low Decrease Soothing, encouraging
4 Low High High Decrease Encouraging
5 High Low Low Const Anger, surprised
6 High Low High Increase Satisfied
7 High High Low Const Encouraging, surprised
8 High High High Decrease Satisfied, surprised
Example 1: a car driving game
The first version of the game does not use any
adaptation methods.
The second version uses a classic method of
dynamic adaptation based on achieved levels of
player results. This changes the game's dynamics
and difficulty by altering environmental features
like fog, rain, darkness, and other factors.
The third version employs a dynamic adaptation
method that detects patterns in the player's
learning curve.
Example 2:
NPC’s with
simplified
representation
of emotional
state
Discussion
• Some players do not manifest their emotion by facial expressions while playing,
while others exaggerate their emotions in order to manipulate the game control
and to obtain a desired difficulty level by cheating the adaptation controller
• Measuring physiological signals by hardware devices and sensors is not
appropriate neither for online games (played at any place and time) nor for
playing desktop or console games in mass, even the sensors communicate
measured data in a wireless way.
• For user-centric adaptation, we could apply individual engagement, attention, or
motivation, provided we could measure or estimate these metrics.
• Measuring engagement and attention could be problematic, hence, classification
or clusterization approaches over physiological data can be applied here.
Conclusions
• The emotion-based adaptation approach is very general and could be
applied for tailoring other features of any serious video game [29] such as:
• Game mechanics – especially game rules, interface, and task automation;
• Informative feedback and help for each game task;
• Educational content;
• Audio-visual effects – such as both the tempo and volume of the music,
illuminations and shadowing, etc.
• This approach offers more engaging and tailored gameplay experiences
that cater to individual players' abilities and learning progress.
Thank you for your attention!
For questions – email to:
madvojd@gmail.com

More Related Content

Similar to Intelligent Adaptation of Difficulty and NPC Behavior in Serious Video Games for Learning

New Microsoft PowerPoint Presentation.pptx
New Microsoft PowerPoint Presentation.pptxNew Microsoft PowerPoint Presentation.pptx
New Microsoft PowerPoint Presentation.pptx
nilesh405711
 
Playability and Player Experience Research
Playability and Player Experience ResearchPlayability and Player Experience Research
Playability and Player Experience Research
ナム-Nam Nguyễn
 
Tobi Saulnier - Cognitive Bias Training Game Valuable for Everything from Law...
Tobi Saulnier - Cognitive Bias Training Game Valuable for Everything from Law...Tobi Saulnier - Cognitive Bias Training Game Valuable for Everything from Law...
Tobi Saulnier - Cognitive Bias Training Game Valuable for Everything from Law...
SeriousGamesAssoc
 

Similar to Intelligent Adaptation of Difficulty and NPC Behavior in Serious Video Games for Learning (20)

Presentation nanjing 2016
Presentation nanjing 2016Presentation nanjing 2016
Presentation nanjing 2016
 
Th202 slides
Th202 slidesTh202 slides
Th202 slides
 
Learning Analytics Serious Games Cognitive Disabilities
Learning Analytics Serious Games Cognitive DisabilitiesLearning Analytics Serious Games Cognitive Disabilities
Learning Analytics Serious Games Cognitive Disabilities
 
Game tech 2014 cognition and game design designing for cognitive adaptabilit...
Game tech 2014 cognition and game design  designing for cognitive adaptabilit...Game tech 2014 cognition and game design  designing for cognitive adaptabilit...
Game tech 2014 cognition and game design designing for cognitive adaptabilit...
 
Gamification
GamificationGamification
Gamification
 
Serious games: current uses and emergent trends
Serious games: current uses and emergent trends Serious games: current uses and emergent trends
Serious games: current uses and emergent trends
 
GAMES USER RESEARCH: Guest Lecture in UX Design Class at Wilfried Laurier Uni...
GAMES USER RESEARCH: Guest Lecture in UX Design Class at Wilfried Laurier Uni...GAMES USER RESEARCH: Guest Lecture in UX Design Class at Wilfried Laurier Uni...
GAMES USER RESEARCH: Guest Lecture in UX Design Class at Wilfried Laurier Uni...
 
Aft 157 design process project -iii
Aft 157 design process project -iiiAft 157 design process project -iii
Aft 157 design process project -iii
 
Games For Upper-limb Stroke Rehabilitation (Seminar)
Games For Upper-limb Stroke Rehabilitation (Seminar)Games For Upper-limb Stroke Rehabilitation (Seminar)
Games For Upper-limb Stroke Rehabilitation (Seminar)
 
Thesis review Presentation
Thesis review PresentationThesis review Presentation
Thesis review Presentation
 
Development of video-based emotion recognition using deep learning with Googl...
Development of video-based emotion recognition using deep learning with Googl...Development of video-based emotion recognition using deep learning with Googl...
Development of video-based emotion recognition using deep learning with Googl...
 
New Microsoft PowerPoint Presentation.pptx
New Microsoft PowerPoint Presentation.pptxNew Microsoft PowerPoint Presentation.pptx
New Microsoft PowerPoint Presentation.pptx
 
Playability and Player Experience Research
Playability and Player Experience ResearchPlayability and Player Experience Research
Playability and Player Experience Research
 
Tobi Saulnier - Cognitive Bias Training Game Valuable for Everything from Law...
Tobi Saulnier - Cognitive Bias Training Game Valuable for Everything from Law...Tobi Saulnier - Cognitive Bias Training Game Valuable for Everything from Law...
Tobi Saulnier - Cognitive Bias Training Game Valuable for Everything from Law...
 
98 374 Lesson 04-slides
98 374 Lesson 04-slides98 374 Lesson 04-slides
98 374 Lesson 04-slides
 
SDM_Game development methodology
SDM_Game development methodologySDM_Game development methodology
SDM_Game development methodology
 
Quiana bradshaw final defense slides
Quiana bradshaw final defense slidesQuiana bradshaw final defense slides
Quiana bradshaw final defense slides
 
Video Games for Education by ADAPTIMES (ADAPTIve player-centric serious vid...
Video Games for Education by ADAPTIMES (ADAPTIve player-centric serious vid...Video Games for Education by ADAPTIMES (ADAPTIve player-centric serious vid...
Video Games for Education by ADAPTIMES (ADAPTIve player-centric serious vid...
 
10 Reasons Why Data-driven App Design Needs Social Science | Julian Runge
10 Reasons Why Data-driven App Design Needs Social Science | Julian Runge10 Reasons Why Data-driven App Design Needs Social Science | Julian Runge
10 Reasons Why Data-driven App Design Needs Social Science | Julian Runge
 
Skill Mapping
Skill MappingSkill Mapping
Skill Mapping
 

More from bbontchev

Towards an Automatized Construction of Personalizable Video Games for Education
Towards an Automatized Construction  of Personalizable Video Games for EducationTowards an Automatized Construction  of Personalizable Video Games for Education
Towards an Automatized Construction of Personalizable Video Games for Education
bbontchev
 

More from bbontchev (8)

Intelligent Approaches for Tailoring of Video Games
Intelligent Approaches for Tailoring of Video GamesIntelligent Approaches for Tailoring of Video Games
Intelligent Approaches for Tailoring of Video Games
 
Генериране на сериозни видео игри-лабиринти
Генериране на сериозни видео игри-лабиринтиГенериране на сериозни видео игри-лабиринти
Генериране на сериозни видео игри-лабиринти
 
Генериране на сериозни видео игри с математически задачи
Генериране на сериозни видео игри с математически задачиГенериране на сериозни видео игри с математически задачи
Генериране на сериозни видео игри с математически задачи
 
Controlling Adaptation in Affective Serious Games
Controlling Adaptation in Affective Serious GamesControlling Adaptation in Affective Serious Games
Controlling Adaptation in Affective Serious Games
 
Учебни видео игри – как да играем
Учебни видео игри – как да играемУчебни видео игри – как да играем
Учебни видео игри – как да играем
 
Учебни видео игри и стилове на учене и на играене
Учебни видео игри и стилове на учене и на играенеУчебни видео игри и стилове на учене и на играене
Учебни видео игри и стилове на учене и на играене
 
Резултати от изпълнение на проект АПОГЕЙ (ДН 12/7 от 15.12.2017г.)
Резултати от изпълнение на проект АПОГЕЙ (ДН 12/7 от 15.12.2017г.)Резултати от изпълнение на проект АПОГЕЙ (ДН 12/7 от 15.12.2017г.)
Резултати от изпълнение на проект АПОГЕЙ (ДН 12/7 от 15.12.2017г.)
 
Towards an Automatized Construction of Personalizable Video Games for Education
Towards an Automatized Construction  of Personalizable Video Games for EducationTowards an Automatized Construction  of Personalizable Video Games for Education
Towards an Automatized Construction of Personalizable Video Games for Education
 

Recently uploaded

+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
Health
 
TECUNIQUE: Success Stories: IT Service provider
TECUNIQUE: Success Stories: IT Service providerTECUNIQUE: Success Stories: IT Service provider
TECUNIQUE: Success Stories: IT Service provider
mohitmore19
 
AI Mastery 201: Elevating Your Workflow with Advanced LLM Techniques
AI Mastery 201: Elevating Your Workflow with Advanced LLM TechniquesAI Mastery 201: Elevating Your Workflow with Advanced LLM Techniques
AI Mastery 201: Elevating Your Workflow with Advanced LLM Techniques
VictorSzoltysek
 
introduction-to-automotive Andoid os-csimmonds-ndctechtown-2021.pdf
introduction-to-automotive Andoid os-csimmonds-ndctechtown-2021.pdfintroduction-to-automotive Andoid os-csimmonds-ndctechtown-2021.pdf
introduction-to-automotive Andoid os-csimmonds-ndctechtown-2021.pdf
VishalKumarJha10
 

Recently uploaded (20)

Unlocking the Future of AI Agents with Large Language Models
Unlocking the Future of AI Agents with Large Language ModelsUnlocking the Future of AI Agents with Large Language Models
Unlocking the Future of AI Agents with Large Language Models
 
How To Use Server-Side Rendering with Nuxt.js
How To Use Server-Side Rendering with Nuxt.jsHow To Use Server-Side Rendering with Nuxt.js
How To Use Server-Side Rendering with Nuxt.js
 
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdf
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdfLearn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdf
Learn the Fundamentals of XCUITest Framework_ A Beginner's Guide.pdf
 
VTU technical seminar 8Th Sem on Scikit-learn
VTU technical seminar 8Th Sem on Scikit-learnVTU technical seminar 8Th Sem on Scikit-learn
VTU technical seminar 8Th Sem on Scikit-learn
 
Azure_Native_Qumulo_High_Performance_Compute_Benchmarks.pdf
Azure_Native_Qumulo_High_Performance_Compute_Benchmarks.pdfAzure_Native_Qumulo_High_Performance_Compute_Benchmarks.pdf
Azure_Native_Qumulo_High_Performance_Compute_Benchmarks.pdf
 
A Secure and Reliable Document Management System is Essential.docx
A Secure and Reliable Document Management System is Essential.docxA Secure and Reliable Document Management System is Essential.docx
A Secure and Reliable Document Management System is Essential.docx
 
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
+971565801893>>SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHAB...
 
Right Money Management App For Your Financial Goals
Right Money Management App For Your Financial GoalsRight Money Management App For Your Financial Goals
Right Money Management App For Your Financial Goals
 
call girls in Vaishali (Ghaziabad) 🔝 >༒8448380779 🔝 genuine Escort Service 🔝✔️✔️
call girls in Vaishali (Ghaziabad) 🔝 >༒8448380779 🔝 genuine Escort Service 🔝✔️✔️call girls in Vaishali (Ghaziabad) 🔝 >༒8448380779 🔝 genuine Escort Service 🔝✔️✔️
call girls in Vaishali (Ghaziabad) 🔝 >༒8448380779 🔝 genuine Escort Service 🔝✔️✔️
 
TECUNIQUE: Success Stories: IT Service provider
TECUNIQUE: Success Stories: IT Service providerTECUNIQUE: Success Stories: IT Service provider
TECUNIQUE: Success Stories: IT Service provider
 
The Ultimate Test Automation Guide_ Best Practices and Tips.pdf
The Ultimate Test Automation Guide_ Best Practices and Tips.pdfThe Ultimate Test Automation Guide_ Best Practices and Tips.pdf
The Ultimate Test Automation Guide_ Best Practices and Tips.pdf
 
AI Mastery 201: Elevating Your Workflow with Advanced LLM Techniques
AI Mastery 201: Elevating Your Workflow with Advanced LLM TechniquesAI Mastery 201: Elevating Your Workflow with Advanced LLM Techniques
AI Mastery 201: Elevating Your Workflow with Advanced LLM Techniques
 
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...
The Real-World Challenges of Medical Device Cybersecurity- Mitigating Vulnera...
 
How To Troubleshoot Collaboration Apps for the Modern Connected Worker
How To Troubleshoot Collaboration Apps for the Modern Connected WorkerHow To Troubleshoot Collaboration Apps for the Modern Connected Worker
How To Troubleshoot Collaboration Apps for the Modern Connected Worker
 
HR Software Buyers Guide in 2024 - HRSoftware.com
HR Software Buyers Guide in 2024 - HRSoftware.comHR Software Buyers Guide in 2024 - HRSoftware.com
HR Software Buyers Guide in 2024 - HRSoftware.com
 
Introducing Microsoft’s new Enterprise Work Management (EWM) Solution
Introducing Microsoft’s new Enterprise Work Management (EWM) SolutionIntroducing Microsoft’s new Enterprise Work Management (EWM) Solution
Introducing Microsoft’s new Enterprise Work Management (EWM) Solution
 
Tech Tuesday-Harness the Power of Effective Resource Planning with OnePlan’s ...
Tech Tuesday-Harness the Power of Effective Resource Planning with OnePlan’s ...Tech Tuesday-Harness the Power of Effective Resource Planning with OnePlan’s ...
Tech Tuesday-Harness the Power of Effective Resource Planning with OnePlan’s ...
 
8257 interfacing 2 in microprocessor for btech students
8257 interfacing 2 in microprocessor for btech students8257 interfacing 2 in microprocessor for btech students
8257 interfacing 2 in microprocessor for btech students
 
AI & Machine Learning Presentation Template
AI & Machine Learning Presentation TemplateAI & Machine Learning Presentation Template
AI & Machine Learning Presentation Template
 
introduction-to-automotive Andoid os-csimmonds-ndctechtown-2021.pdf
introduction-to-automotive Andoid os-csimmonds-ndctechtown-2021.pdfintroduction-to-automotive Andoid os-csimmonds-ndctechtown-2021.pdf
introduction-to-automotive Andoid os-csimmonds-ndctechtown-2021.pdf
 

Intelligent Adaptation of Difficulty and NPC Behavior in Serious Video Games for Learning

  • 1. Intelligent Adaptation of Difficulty and NPC Behavior in Serious Video Games for Learning Ivan Naydenov, Ilko Adamov, Boyan Bontchev SUMMIT Annual Conference Sofia, Bulgaria April 24, 2024 Contract BG-RRP-2.004-0008 for the financing of project "Sofia University - Marking Momentum for Innovation and Technological Transfer“ under pillar 2 "Establishing a network of research higher education institutions in Bulgaria", component "Innovative Bulgaria" from National Recovery and Resilience Plan part of the program to accelerate economic recovery and transformation through science and innovation
  • 2. Two key issues of user-centric adaptation of video games for learning • Tailoring the learning task difficulty – it is very important because didactic tasks are mapped to gaming tasks; • Adapting the behavior of non-player characters (NPCs), i.e., virtual heroes – applied in educational games for: • (a) teaching assistants (the most popular case); • (b) concurrent learners; or • (c) opposite characters or enemies.
  • 3. Recognizing emotions from facial expressions • Understanding human behavior: recognizing emotions from facial expressions allows us to gain insight into individual and collective emotional states. • Communication and social interaction: facial expressions are a fundamental aspect of nonverbal communication. • User experience and human-computer interaction: recognizing emotions from facial expressions is particularly relevant in the field of human-computer interaction – systems and interfaces can be designed to adapt and respond accordingly. • Psychological and mental health assessment: the ability to recognize and measure emotions from facial expressions can be valuable in psychological and mental health assessment.
  • 4. Popular methods for emotion analysis A. By using facial expressions: • Classification using machine learning models like KNN, SVM, etc. • Convolutional Neural Networks (CNNs) B. Measurement and analysis of physiological data like Electrocardiograph (ECG), Blood Volume Pulse (BVP), Galvanic Skin Response (GSR) and Electromyography (EMG) • Measure and extract physiological data • Analyze and recognize emotions by data clustering
  • 5. Recognition of player emotional state by CNN • Colab (https://colab.research.google.com) with various Python libraries for data analysis and visualization • The neural network architecture consists of 779,718 neurons organized into 17 layers divided into four blocks. • Each of the first three blocks contains three convolutional layers and one pooling layer. • Convolution layers employ filters or kernels to extract local features, such as edges, corners, and textures, from the input image. • Pooling layers are applied after convolution to reduce the spatial dimensions of the feature maps using techniques like max pooling or average pooling. • The final block includes a fully connected layer, which combines the high-level features extracted from previous layers to produce the final classification output.
  • 6.
  • 7. Training and validation accuracy (left) and loss (right) of the neural network
  • 9. Samples of BVP (left) and GSR (right) data – used for data clusterization
  • 10. K-Means (left), mean-shift (middle), and agglomerative (right) clustering based on physiological data
  • 11. Adaptation of difficulty and NPC tutor behavior The player: • In situation 1 – has played a lot without any engagement and motivation; • In situation 2 – has achieved high score and demonstrated good skills but remains in apathy; • In situation 3 – is disappointed by his/her low outcome; • In situation 4 – manages to play well but at the price of some non-desired emotions; • In situation 5 – is happy to play the game but without trying to achieve good outcomes; • In situation 6 – is happy to play the educational game while achieving a good score; • In situation 7 – cannot achieve a good score and has both desired and non-desired emotions; • In situation 8 – has succeeded in the game but at the price of negative emotions. Situa- tion Desired Emotion Joy, Anger Non-desired emotion Fear, Sadness Out- comes Difficulty NPC tutor behavior 1 Low Low Low Const Encouraging 2 Low Low High Increase Satisfied, encouraging 3 Low High Low Decrease Soothing, encouraging 4 Low High High Decrease Encouraging 5 High Low Low Const Anger, surprised 6 High Low High Increase Satisfied 7 High High Low Const Encouraging, surprised 8 High High High Decrease Satisfied, surprised
  • 12. Example 1: a car driving game The first version of the game does not use any adaptation methods. The second version uses a classic method of dynamic adaptation based on achieved levels of player results. This changes the game's dynamics and difficulty by altering environmental features like fog, rain, darkness, and other factors. The third version employs a dynamic adaptation method that detects patterns in the player's learning curve.
  • 14. Discussion • Some players do not manifest their emotion by facial expressions while playing, while others exaggerate their emotions in order to manipulate the game control and to obtain a desired difficulty level by cheating the adaptation controller • Measuring physiological signals by hardware devices and sensors is not appropriate neither for online games (played at any place and time) nor for playing desktop or console games in mass, even the sensors communicate measured data in a wireless way. • For user-centric adaptation, we could apply individual engagement, attention, or motivation, provided we could measure or estimate these metrics. • Measuring engagement and attention could be problematic, hence, classification or clusterization approaches over physiological data can be applied here.
  • 15. Conclusions • The emotion-based adaptation approach is very general and could be applied for tailoring other features of any serious video game [29] such as: • Game mechanics – especially game rules, interface, and task automation; • Informative feedback and help for each game task; • Educational content; • Audio-visual effects – such as both the tempo and volume of the music, illuminations and shadowing, etc. • This approach offers more engaging and tailored gameplay experiences that cater to individual players' abilities and learning progress.
  • 16. Thank you for your attention! For questions – email to: madvojd@gmail.com