Game Analytics Project


This project is conducted by the Knowledge Discovery and Machine Learning group of the Graduate Program in Informatics at Pontifícia Universidade Católica do Paraná.

Objective:

To better comprehend players through propositions of useful metrics extracted from telemetry data, and identify and mitigate risk situations of games usage lifecycles.
In special, we focus on the following subjects:
- Players' motivations
- Players' approval over game contents
- Players' commitment
- Players' social interactions
- Content consumption

The generated metrics can be applied to deal with different risk situations, such as the identification of churn (from a proactive or reactive perspective) and survival time, community management, player simulation, approval of game upgrades, suggestions of suitable new game contents, the identification of monetization candidates, among others.

Approaches

Commitment

The commitment metric carries the idea of identifying players' degrees of engagement, usually dividing them into three groups, low, average, and high. This metric is attached to the amount of played time and obtained score of players, which means that the more remarkable are those characteristics, the more committed a player is.

Psychological profile

The proposed psychological profile of players is a set of metrics conceived based on an analysis of over 67 textual psychological models, comprehending short, mid, and long-term aspects of human nature. This psychological profile is obtained from the players' actions in-game, and its identification process was coined as "Data Enhancement", portraying the players' motivations and their success or not in attaining them in an individualized manner.

Social interactions

Socialization is known as one of the most relevant motivational aspects to play games; thus, we wish to give special attention to this topic by providing metrics that depicts the social interactions (e.g., influence, popularity, group structure, etc.).

Published Papers

KUMMER, L. B. M.; BINDER, F.; BARBON JR, S.; NIEVOLA, J. C.; PARAISO, E. C.
Predição do Estágio de Nicho em Jogos RPG Massivos de Multijogadores utilizando o Comprometimento (in Portuguese) In: Encontro Nacional de Inteligência Artificial e Computacional (ENIAC), 2016, Recife. Encontro Nacional de Inteligência Artificial e Computacional (ENIAC)., 2016. p.289 - 300

KUMMER, L. B. M.; NIEVOLA, J. C.; PARAISO, E. C.
Digital Game Usage Lifecycle: a systematic literature review In: Simpósio Brasileiro de Games (SBGames), 2017, Curitiba. 2017 16th Brazilian Symposium on Computer Games and Digital Entertainment (SBGames), 2017.

KUMMER, L. B. M.; NIEVOLA, J. C.; PARAISO, E. C.
A Key Risk Indicator for the Game Usage Lifecycle In: The Florida Artificial Intelligence Research Society (FLAIRS), 2017, Marco Island, Florida, USA.
Proceedings of the Thirtieth International Florida Artificial Intelligence Research Society Conference., 2017. p.394 - 399

KUMMER, L. B. M.; NIEVOLA, J. C.; PARAISO, E. C.
Identifying Niche Stage in MMORPGs Using Ensemble Classifier In: Proceedings of the Thirty-first International Florida Artificial Intelligence Research Society Conference, 2018, Melbourne. The Florida Artificial Intelligence Research Society (FLAIRS)., 2018.

KUMMER, L. B. M.; NIEVOLA, J. C.; PARAISO, E. C.
Applying Commitment to Churn and Remaining Players Lifetime Prediction In: Computational Intelligence and Games (CIG), 2018 IEEE Conference, 2018, Maastricht.
Proceedings-CIG2018. , 2018. p.213 - 220

KUMMER, L. B. M.; IIDA, H.; NIEVOLA, J. C.; PARAISO, E. C.
Identifying Influences of Game Upgrades on Profitable Players Behavior in MMORPGs. In: Entertainment Computing and Serious Games, 2019, Arequipa.
First IFIP TC 14 Joint International Conference, ICEC-JCSG 2019 Arequipa, Peru, November 11–15, 2019 Proceedings. , 2019.

Dissertations and Thesis:

KUMMER, LUIZ BERNARDO MARTINS.
A Method to Predict Risk Situations on Digital Game Usage Lifecycle. 2017. Master's dissertation. Pontifícia Universidade Católica do Paraná.

Patents

Psychological profile identification

Software registration under process.

Databases

Original Databases

- World of Warcraft Avatar History Dataset: http://mmnet.iis.sinica.edu.tw/dl/wowah/
- Blade&Soul (CIG 2017 Data Mining competition): https://datathon.ncsoft.com/2017_competition_data
- League of Legends (data retrieve API): https://developer.riotgames.com/apis

Preprocessed Databases

- Blade&Soul (from paper "Applying Commitment to Churn and Remaining Players Lifetime Prediction"): https://www.ppgia.pucpr.br/~paraiso/Projects/GameAnalytics/DataBases/BladeAndSoul/

- League Of Legends (from dissertation "Measuring the Commitment of MOBA players through their in-game tasks"): https://www.ppgia.pucpr.br/~paraiso/Projects/GameAnalytics/DataBases/LeagueOfLegends2023/

Researchers

Ph.D. Students


Luiz Bernardo Martins Kummer
Email: [email protected]

received a BA in Information Systems and an M.Sc. in Computer Science at Pontificia Universidade Catolica do Parana (PUC-PR). Currently, he is a Ph.D. student at PUC-PR, having studied for one year at the Japan Advanced Institute of Science and Technology (JAIST) as a special visiting student of a split-site doctoral program. His research interest lies at the identification of players' commitment to playing games based on telemetry data, and player simulation.

M.Sc. students


João Felipe Humenhuk
Email: [email protected]

received a BA in Computer Engineering at Universidade Positivo (UP). Currently, he is a M.Sc. student at Pontificia Universidade Catolica do Parana (PUC-PR). His research interest lies in the application of machine learning algorithms into diverse areas like biology, music, game analytics, among others, to solve problems such as churn prediction, human machine interface, music generation, etc.

Core Faculty - PPGIA - PUCPR


Prof. Dr. Julio Cesar Nievola
Email: [email protected]

received the B.S. degree in Electrical Engineering from Universidade Federal Tecnológica do Paraná, Curitiba, Brasil, in 1984 and the M.S. and Ph.D. degrees in Electrical Engineering from Universidade Federal de Santa Catarina, Florianópolis, Brasil, in 1988 and 2005, respectively. He is currently the Leader of the Knowledge Discovery and Machine Learning Group (DCAM) at Programa de Pós-Graduação em Informática – PPGIa, Pontifícia Universidade Católica do Paraná - PUCPR. In 2001 he was a Visiting Professor at Fachhoschule Konstanz, in Germany and in 2008/2009 he was a Fellow Visiting Professor at University of Kent at Canterbury, UK. His main research interest includes data mining, machine learning, time series analysis and bioinformatics.

Prof. Dr. Emerson Cabrera Paraiso
Email: [email protected]

received the M.S. degree in Electrical Engineering and Industrial Informatics at Universidade Tecnológica Federal do Paraná – Brazil and the Ph.D. degree in Information Technology and Systems at Université de Technologie de Compiègne - France. He is a member of the Knowledge Discovery and Machine Learning Research Group at the Graduate Program in Informatics at Pontifícia Universidade Católica do Paraná - PUCPR-Brazil. His main research interests span various issues such as Natural Language Processing, Knowledge Engineering and Text and Data Mining.

Partnerships

Japan Advanced Institute of Science and Technology (JAIST) - Japan

Prof. Dr. Hiroyuki Iida
Email: [email protected]

received his Ph.D. in 1994 on Heuristic Theories on Game-Tree Search from the Tokyo University of Agriculture and Technology, Japan. Since 2005, he has been a Professor at JAIST, where he is also Trustee and Vice President of Educational and Student Affairs. He is the head of the Iida laboratory and has published over 300 hundred papers, presentations, and books. His current research interests include artificial intelligence, game informatics, game theory, mathematical modeling, search algorithms, game-refinement theory, game tree search, and entertainment science.