cover
Contact Name
Andhika Rafi Hananto
Contact Email
andhikarh90@gmail.com
Phone
+62895422720524
Journal Mail Official
support@ijrm.net
Editorial Address
Puri Mersi Baru, Blok A2, Jl. Martadireja 2 Purwokerto, Kab. Banyumas,Jawa Tengah.
Location
Kab. banyumas,
Jawa tengah
INDONESIA
International Journal Research on Metaverse
Published by Meta Bright Indonesia
ISSN : -     EISSN : 30626927     DOI : https://doi.org/10.47738/ijrm
Core Subject : Science,
Virtual and augmented reality technologies Network infrastructure and architecture for the metaverse Digital economy and transactions in the metaverse Social and cultural aspects of virtual environments Development and design of content in the metaverse Impact of the metaverse on industries such as education, healthcare, entertainment, and business Regulation, policy, and ethics in the metaverse IJRM aims to foster interdisciplinary dialogue and collaboration, contributing to the body of knowledge that drives the adoption and evolution of metaverse technologies. Papers published in IJRM are grounded in rigorous research methods and are expected to articulate their implications for theory and practice clearly. Authors are encouraged to state their contributions to the state-of-the-art in the field explicitly. Subject Area and Category: The International Journal Research on Metaverse focuses on virtual and augmented reality, network infrastructure, digital economy, social and cultural impacts, content development, industry-specific applications, regulation and ethics, and practical case studies.
Articles 35 Documents
Predicting FIFA Ultimate Team Player Market Prices: A Regression-Based Analysis Using XGBoost Algorithms from FIFA 16-21 Dataset Warmayana, I Gede Agus Krisna; Yamashita, Yuichiro; Oka, Nobuto
International Journal Research on Metaverse Vol. 2 No. 2 (2025): Regular Issue June 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i2.25

Abstract

This study investigates the use of XGBoost, a machine learning algorithm, for predicting player prices in FIFA Ultimate Team (FUT) from FIFA 16 to FIFA 21. Virtual economies in gaming, particularly in FUT, have grown substantially, with in-game asset prices influenced by a variety of factors such as player attributes, performance metrics, and market dynamics. The objective of this research is to enhance the accuracy of price predictions in FUT through advanced machine learning techniques. The dataset comprises historical player data, including attributes such as rating, skills, and in-game statistics. XGBoost was employed due to its ability to handle large, complex datasets and capture non-linear relationships effectively. The model achieved an R-squared value of 0.8911, indicating that it explains 89% of the variance in player prices, while the RMSE value of 30368.85 reveals the model's precision in estimating prices. Feature importance analysis showed that attributes such as WorkRate and Rating significantly influenced price predictions. Compared to traditional methods like linear regression, XGBoost provided superior accuracy and computational efficiency, making it a valuable tool for understanding player price dynamics in virtual gaming markets. The findings suggest that accurate price predictions can improve gaming strategies for players and provide valuable insights for game developers in optimizing virtual economies. This research also highlights the potential for further exploration using advanced machine learning algorithms to predict price fluctuations in gaming environments.
Analyzing Player Performance Metrics for Rank Prediction in Valorant Using Random Forest: A Data-Driven Approach to Skill Profiling in the Metaverse Rahardja, Untung; Aini, Qurotul
International Journal Research on Metaverse Vol. 2 No. 2 (2025): Regular Issue June 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i2.26

Abstract

This study explores the application of Random Forest, a powerful data mining technique, to predict player ranks in Valorant, a competitive first-person shooter. By analyzing a range of player performance metrics, including headshots, kills, damage received, and traded kills, the study identifies the key features that influence player rank determination. Using a dataset of player statistics, the model was trained to predict player ranks, achieving a prediction accuracy of 50.09%. The analysis revealed that headshots and traded kills were the most influential metrics in determining player rank, suggesting that skill-based metrics like accuracy and tactical gameplay are crucial for ranking in the game. These findings highlight the importance of understanding the relationship between various performance indicators and rank progression, offering valuable insights for both game developers and players. The results contribute to the growing body of research in gaming analytics, showcasing how data mining techniques can be used to analyze player behavior and improve competitive balance in games. The study underscores the potential of using data-driven approaches to enhance game design, providing developers with actionable insights to refine rank prediction systems, adjust in-game mechanics, and ensure a more balanced competitive environment. Looking ahead, future research can explore the use of alternative machine learning models, such as support vector machines (SVM), XGBoost, or neural networks, to improve the prediction accuracy and robustness of the model. Additionally, expanding the dataset to include more detailed player behaviors, match outcomes, and even temporal aspects of player performance could provide a more comprehensive understanding of the factors influencing player ranks. This can help further unravel the complexities of player behavior and performance in the metaverse, where virtual environments evolve dynamically based on player interactions.
Harnessing Sentiment Analysis with VADER for Gaming Insights: Analyzing User Reviews of Call of Duty Mobile through Data Mining Batumalay, Malathy; S, Priya; Kumar, Vinoth
International Journal Research on Metaverse Vol. 2 No. 2 (2025): Regular Issue June 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i2.27

Abstract

This study investigates the application of sentiment analysis to understand user feedback for Call of Duty Mobile, a highly popular mobile game, by analyzing 50,000 reviews sourced from the Google Play Store. The research aimed to extract actionable insights from user sentiments, which could guide future game development and improvement. To achieve this, the sentiment of each review was analyzed using VADER (Valence Aware Dictionary and sEntiment Reasoner), a robust tool for classifying sentiment in textual data. The study categorizes reviews into three sentiment groups—positive, negative, and neutral—to identify and analyze prevailing user emotions. The findings revealed that the majority of reviews were positive, with users primarily praising the gameplay, graphics, and overall mobile experience. These aspects were considered crucial in driving user satisfaction and contributed to a majority of the positive feedback. Conversely, negative reviews were often focused on issues such as network connectivity problems, long loading times, and performance errors, indicating areas where users experienced frustration. These results highlight the importance of technical performance and network stability as key factors influencing player satisfaction. The study also delved deeper into keyword analysis to uncover common themes in the reviews, such as in-app purchases and concerns related to technical performance, which were frequently mentioned by users in both positive and negative feedback. These insights provide developers with a clearer understanding of what players value most in the game and where improvements are necessary. The study concludes that sentiment analysis can serve as a powerful tool for understanding user feedback, offering developers a data-driven approach to enhance game features and address user concerns. Moving forward, future research could benefit from the application of additional machine learning models to refine sentiment classification accuracy, as well as the integration of cross-platform reviews to gain a more comprehensive understanding of player sentiment across different user groups and devices. Such approaches would provide a richer, more nuanced view of user experiences, enabling game developers to create even more engaging and satisfying gaming experiences.
Estimating Player Market Value in Virtual Leagues: A Clustering Approach Using Player Attributes for Metaverse Applications Turistiati, Ade Tuti; Monk, Lincoln James Faikar; Ramadhan, Hafizh Faikar Agung
International Journal Research on Metaverse Vol. 2 No. 2 (2025): Regular Issue June 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i2.28

Abstract

The advent of virtual environments, particularly within the Metaverse, has revolutionized the way sports simulations and virtual leagues operate. In these environments, understanding and predicting player market value is essential for optimizing team management, player scouting, and in-game strategies. This paper presents a clustering approach using K-Means to segment players based on their performance attributes and predict their market value in virtual leagues. The dataset includes various player attributes such as age, goals scored, assists, minutes played, and performance metrics like expected goals (xG) and expected assists (xA). The K-Means clustering algorithm was applied to partition players into three distinct groups based on their performance profiles. The results indicated that high-performing players, characterized by high goals scored, assists, and other key metrics, were grouped in one cluster, while lower-performing players were segmented into another. These clusters correspond to different player market values, with higher-performance clusters being associated with higher market value. The clustering analysis reveals significant patterns that can inform virtual league operations, including player trading, recruitment, and team-building strategies. The findings suggest that virtual league developers, managers, and gamers can leverage these clusters to make more informed decisions regarding player acquisitions and team compositions. Furthermore, the clustering results can be used to dynamically adjust player values based on their performance attributes, offering a realistic simulation of real-world sports economics. Future research may explore more advanced clustering techniques, such as hierarchical clustering, and expand the dataset to include additional attributes like player psychology or external factors like fan sentiment. Overall, this paper highlights the potential of clustering algorithms to enhance player market valuation and decision-making within virtual leagues.
Analyzing the Impact of Social Media and Influencer Endorsements on Game Revenue using Multiple Linear Regression in the Metaverse Dewi, Deshinta Arrova; Kurniawan, Tri Basuki
International Journal Research on Metaverse Vol. 2 No. 2 (2025): Regular Issue June 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i2.29

Abstract

The gaming industry, particularly within the metaverse, has seen significant transformations driven by the integration of social media, influencer marketing, and player engagement metrics. These elements are crucial in shaping the success and revenue generation of games. This study explores the role of social media mentions and influencer endorsements in influencing game revenue, applying DBSCAN clustering to segment player engagement into distinct groups. By analyzing the "Gaming Trend 2024" dataset, which includes key metrics such as social media mentions, influencer endorsements, in-game purchases, and game revenue, we identify patterns in player behavior that directly impact revenue generation. The DBSCAN clustering method was employed to group players based on their social media interactions and influencer influence, identifying key segments that contribute to game success. The results reveal that certain clusters, characterized by higher social media engagement and influencer endorsements, are associated with increased game revenue. In contrast, other segments showed lower engagement and contributed less to overall revenue. The clustering analysis highlights the power of social media and influencers in driving player behavior, which in turn drives financial outcomes for game developers. This research provides insights into how targeted marketing strategies, personalized influencer campaigns, and tailored engagement efforts can enhance game revenue. This study offers practical applications for game developers and marketers in the metaverse, emphasizing the need to leverage clustering insights to optimize marketing strategies and increase revenue. Future research could expand on these findings by integrating sentiment analysis of social media mentions, exploring alternative clustering methods like hierarchical clustering, and developing hybrid models that combine clustering with predictive analytics to forecast game revenue trends.
Classification of Game Genres Based on Interaction Patterns and Popularity in the Virtual World of Roblox Hasanah, Uswatun; Sunarko, Budi; Hidayat, Syahroni; Rachmawati, Rina
International Journal Research on Metaverse Vol. 2 No. 3 (2025): Regular Issue September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i3.30

Abstract

The rapid growth of user-generated virtual environments has elevated the importance of understanding player behavior and content dynamics in metaverse platforms. This study investigates the relationship between game genres and user engagement in Roblox, one of the largest and most interactive virtual worlds. Utilizing a dataset of over 300 game entries, we analyzed engagement metrics including visits (ranging from thousands to over 2.8 billion), likes (up to 1,000,000), favorites (up to 3.4 million), and active user counts (as high as 22,155). Descriptive statistics and correlation analysis revealed that action-oriented genres—particularly Action, Shopping, and Obby & Platformer—consistently outperform others in attracting and retaining users. The strong positive correlation between likes and favorites (r = 0.95) indicates that user satisfaction strongly predicts long-term interest, while negative feedback (dislikes) shows minimal correlation with other variables. In contrast, genres such as Education and Entertainment demonstrated significantly lower averages, with visits below 1 million, and active user counts typically under 1,000. These findings provide practical insights for developers and platform administrators seeking to optimize content strategies and offer a foundation for future research involving clustering analysis, sentiment mining, and temporal behavior modeling to enhance recommendation systems and genre personalization within metaverse ecosystems.
Temporal Analysis of Blockchain Transactions in the Metaverse Using Time Series Guballo, Jayvie Ochona
International Journal Research on Metaverse Vol. 2 No. 3 (2025): Regular Issue September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i3.32

Abstract

This study aims to analyze the temporal data of blockchain transactions in the metaverse using time series analysis techniques such as ARIMA and LSTM. The primary focus of this research is to identify significant trends and time patterns in transaction activities within the metaverse. By employing ARIMA, the time series data is decomposed into trend, seasonal, and residual components, providing crucial insights into its structure. The ARIMA model demonstrated a mean absolute error (MAE) of 10,525.73, a mean squared error (MSE) of 150,247,506.45, and a root mean squared error (RMSE) of 12,259.65, indicating a reasonably good fit with some potential for improvement. To capture more complex temporal dependencies in the data, an LSTM model was also applied. The performance of the LSTM model, evaluated using RMSE, was 10.0 for the training set and 15.0 for the testing set. The higher RMSE on the testing set indicates slight overfitting, where the model fits the training data better than unseen data. However, the LSTM model showed strong capability in predicting daily transaction values with fairly high accuracy, despite some minor discrepancies between actual and predicted values. Descriptive statistical analysis of the transaction data revealed that the average daily transaction volume was 108,225.72 with a standard deviation of 8,489.47, indicating significant variability. The daily transaction range spanned from 83,052.86 to 134,869.80, reflecting a wide variation in transaction volume. The results of this study highlight the importance of temporal analysis in understanding blockchain transactions in the metaverse. Insights gained from this analysis can assist in strategic planning and decision-making within the metaverse ecosystem. By further refining model tuning and employing more advanced analysis techniques, predictive accuracy can be enhanced, providing more comprehensive insights and more accurate predictions of transaction behavior.
Data-Driven Imagination Genre Clustering of Anime Content to Inspire Culturally Rich Metaverse Spaces Chantanasut, Thaworada
International Journal Research on Metaverse Vol. 2 No. 3 (2025): Regular Issue September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i3.33

Abstract

The growing demand for culturally immersive experiences in virtual environments has highlighted the importance of integrating narrative-driven content into metaverse design. This study applies a data-driven clustering approach to categorize 2,000 anime titles based on genres, themes, audience demographics, and user engagement metrics such as score, number of users, and member count. Using K-Means clustering and Principal Component Analysis (PCA), five distinct clusters were identified, each reflecting unique narrative typologies and audience preferences. The resulting clusters reveal meaningful thematic patterns: Cluster 0 emphasizes action and adventure with an average score of 8.56 and over 1 million members; Cluster 1 is centered around fantasy and supernatural elements with a dominant Shounen demographic; Cluster 2 comprises psychological and sci-fi anime with high intellectual engagement; Cluster 3 features emotionally resonant titles like romance and slice of life with the highest average score of 8.78; and Cluster 4 presents genre-diverse content with a focus on comedy and school life. PCA visualization confirmed the coherence of these groupings in two-dimensional space, and genre frequency analysis showed that Action, Comedy, and Drama were the most prevalent across the dataset. The findings offer actionable insights for culturally intelligent metaverse development, proposing each genre cluster as a thematic blueprint for designing distinct virtual environments. These results demonstrate how narrative clustering can bridge media analytics with user-centered virtual worldbuilding.
Temporal Analysis of Viewer Engagement in One Piece: Trends in IMDb Ratings Across Arcs and Time Selvaraj, Poovarasan; Yang, Qingxue
International Journal Research on Metaverse Vol. 2 No. 3 (2025): Regular Issue September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i3.34

Abstract

This study examines the long-term viewer reception of One Piece, a globally influential anime series, by analyzing 1,122 episodes released between 1999 and 2023 using IMDb data. The research aims to understand how audience engagement and critical reception evolve, and how content structure, specifically narrative arcs and episode types, affects these dynamics. Quantitative analysis was conducted on episode ratings, vote counts, narrative arc segmentation, and episode classification (Canon, Filler, Semi-Filler). The findings reveal a clear upward trend in average episode ratings, increasing from 7.9 in 1999 to over 8.4 in recent years. Viewer participation also grew significantly, with total annual votes rising from 63,939 in 1999 to over 300,000 votes per year after 2020. Canon episodes achieved the highest average rating (8.22) across 996 episodes and garnered a total of 2,630,987 votes. In contrast, Filler episodes averaged 6.62 (90 episodes), while Semi-Filler episodes scored 7.14 (36 episodes). The highest-rated narrative arcs, each comprising at least five episodes, consistently achieved an average rating of 8.7 or higher and demonstrated alignment with major plot developments and emotional climaxes. These results highlight the importance of narrative relevance and continuity in maintaining audience satisfaction and fostering long-term engagement. The increasing vote volume reflects the expansion of One Piece’s global audience and the role of digital platforms in amplifying participatory behavior. This study highlights how serialized media can maintain cultural impact through strategic narrative design and evolving viewer engagement.
User Transaction Patterns in Smart Contracts Based on Call Frequency and Transfer Value Hery, Hery; Haryani, Calandra
International Journal Research on Metaverse Vol. 2 No. 3 (2025): Regular Issue September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/ijrm.v2i3.35

Abstract

Smart contracts are integral to blockchain technology, enabling decentralized and automated transactions. This study examines 1,000 smart contracts by analyzing metrics such as total transactions, unique users, total value transferred (ETH), gas consumption, and call frequency. Total transactions range from 1 to 18,902, with unique users spanning 1 to 14,839. The average total value transferred is 3,245.87 ETH, peaking at 7,850.16 ETH, while gas consumption averages 25,486,392 units with a maximum of 58,471,065 units. Strong correlations were identified between transaction volume (r = 0.78), user engagement, and gas consumption. Clustering analysis categorizes contracts into low, moderate, and high-activity groups, while anomaly detection highlights 32 contracts with unusual behaviors, indicating inefficiencies or vulnerabilities. These findings emphasize the importance of optimizing smart contract designs to improve efficiency, security, and scalability. The study provides actionable insights into operational patterns and proposes future research directions, including design optimization, real-time monitoring, cross-platform analysis, and machine learning applications for predictive modeling. By addressing these aspects, this research contributes to the ongoing development of robust and efficient decentralized systems.

Page 3 of 4 | Total Record : 35