E.g. there are five different functions a user can ascribe to a badge (Anton & Churchill). ⢠Personalization has been successful in other digital contexts ...
Profile-Based Algorithm for Personalized Gamification in Computer-Supported Collaborative Learning Environments Antti Knutas, Rob van Roy, Timo Hynninen, Marco Granato, Jussi Kasurinen, Jouni Ikonen Lappeenranta University of Tech. & LERO, Irish Software Research Centre KU Leuven University of Milan South-Eastern Finland University of Applied Sciences
Terms and concepts Research questions Need for personalization Overall approach Relevant design theories Rule design structure Outcomes Conclusions
Terms and concepts • Collaborative learning • “students working towards a shared goal with a teacher as a facilitator” • Gamification • ”applying game mechanics to non-game environments” • Self-determination theory • One explanation for why people act (voluntarily) as they do, by Deci & Ryan
Introduction and research questions Motivation -> Gamification, a one size fits all solution? 1. How can personalized gamification features be designed to address the preferences of different user types? 2. How could customized, profile-based gamification challenges be assigned to different users in CSCL environments?
Personalization -> effectiveness? • Different users interpret, functionalize and evaluate the same game elements in radically different way (Koster) • E.g. there are five different functions a user can ascribe to a badge (Anton & Churchill) • Personalization has been successful in other digital contexts
Approach • Deterding’s gamification design process • Synthesis: Apply relevant theories • Self-determination theory + • Design heuristics for effective gamification (van Roy et al.) • Ideation: How to personalize? • Marczewski’s gamification user types + • Lens of intrinsic skill atoms (Deterding) • Iterative prototyping: Rules -> CN2-based rule generator based on expert panel created examples
Design heuristics for effective gamification (van Roy & Deterding; relevant examples) • • • • •
#1 Avoid obligatory uses. #2 Provide a moderate amount of meaningful options. #5 Facilitate social interaction. #7 Align gamification with the goal of the activity in question. #8 Create a need-supporting context.
Marczewski’s1 gamification user type hexad
1. Marczewski, A. (2015). User Types. In Even Ninja Monkeys Like to Play: Gamification, Game Thinking and Motivational Design (1st ed., pp. 65-80). CreateSpace Independent Publishing Platform.
Constructing the rules (an example) • •
Goal: Action:
• • • • •
Object: Rules: Feedback: Challenge: Motivation:
Get other team to assist yours a) Point out a task to the other team b) Task is solved (system state) (system functionality) Notifications, team status (inherent difficulty) Relatedness
Algorithm and system architecture Backend: CN2 rule inducer 1. Interaction
4. Response and gamification tasks
(2). User behavior parameters
(3). Gamification task proposal, if conditions match
Example CN2 rule: IF Hexad = Free Spirit AND Chat Activity != Low AND Ownteam opentasks = high AND Own- team task age = high AND Ownteamactivity != high THEN Challenge_class = 7
Application environment #1
Application environment #2
Conclusions •
We presented an approach to create personalized gamification rulesets using a framework for effective gamification (Goal 1). • This ruleset algorithm can be used as a plugin in computer-supported collaborative learning environments (Goal 2) • Novel results • Personalization through adaptation (one of the first implementations for gamification) • Separating rules from presentation
Bonus slide: All material available libre https://github.com/aknutas/ludusengine