SHAP-Interpretable Cross-Platform Machine Learning for Classifying High-Performing Influencers Using Ratio-Based Engagement Metrics

Muchammad Luqmanul Hakim; Riska Dhenabayu; Ika Diyah Candra Arifah; Anita Safitri

doi:10.57152/malcom.v6i4.2887

Authors

Muchammad Luqmanul Hakim University of Surabaya
Riska Dhenabayu University of Surabaya https://orcid.org/0000-0002-6530-2099
Ika Diyah Candra Arifah University of Surabaya
Anita Safitri University of Surabaya

DOI:

https://doi.org/10.57152/malcom.v6i4.2887

Keywords:

CatBoost, Cross-Platform Classification, High-Performing Influencer, Ratio-Based Engagement Metrics, SHAP

Abstract

This study aims to develop and evaluate a Shapley Additive exPlanations (SHAP)-interpretable cross-platform machine learning framework for classifying high-performing influencer segments on TikTok and YouTube using ratio-based engagement metrics. The study addresses limitations of follower or subscriber-based evaluation by constructing a platform-stratified performance label and applying leakage-controlled feature selection to reduce circular prediction bias. A total of 2,000 influencer records, consisting of 1,000 TikTok and 1,000 YouTube influencers, were analyzed using Random Forest, Extreme Gradient Boosting (XGBoost), Categorical Boosting (CatBoost), and Support Vector Machine (SVM). Model performance was evaluated using accuracy, macro precision, macro recall, macro F1-score, and Receiver Operating Characteristic–Area Under the Curve (ROC-AUC), while SHAP was used to interpret feature contributions. The tuned CatBoost model produced the highest observed performance, with 0.9925 accuracy, 0.9904 macro F1-score, and 0.9995 ROC-AUC. Additional validation showed stable model ranking across alternative thresholds, while proxy leakage ablation and cross-platform transfer indicated that results should be interpreted as internal classification consistency rather than full platform-independent generalization. SHAP analysis showed that ratio-based engagement variables, particularly like ratio and engagement rate, contributed more strongly than raw audience size. The findings demonstrate that influencer performance can be represented as an interpretable

Downloads

Download data is not yet available.

References

Y.-H. Hsiao and Y.-Y. Lin, “Decoding influencer marketing effectiveness on instagram: Insights from image, text, and influencer features,” Journal of Retailing and Consumer Services, vol. 85, p. 104285, 2025, doi: 10.1016/j.jretconser.2025.104285.

Y. Peng and L. Lu, “Untangling influence: The effect of follower-followee comparison on social media engagement,” Journal of Retailing and Consumer Services, vol. 78, p. 103747, 2024, doi: 10.1016/j.jretconser.2024.103747.

Y.-H. Hsiao and Y.-Y. Lin, “Decoding influencer marketing effectiveness on instagram: Insights from image, text, and influencer features,” Journal of Retailing and Consumer Services, vol. 85, p. 104285, Jul. 2025, doi: 10.1016/j.jretconser.2025.104285.

L. Ji, X. Chen, L. Jiang, and J. T. Huang, “Within-and cross-group spillover effects in influencer marketing: The heterogeneity between micro-and macro-influencers,” Information & Management, vol. 62, no. 7, p. 104202, 2025, doi: 10.1016/j.im.2025.104202.

A. Shmuel, O. Glickman, and T. Lazebnik, “A comprehensive benchmark of machine and deep learning models on structured data for regression and classification,” Neurocomputing, p. 131337, 2025, doi: 10.1016/j.neucom.2025.131337.

F. Iqbal, “A comprehensive explainable machine learning framework for predicting sheep body weight,” Small Ruminant Research, p. 107776, 2026, doi: 10.1016/j.smallrumres.2026.107776.

D. M. Utami, M. Ikhsan, T. Dartanto, and R. Mallarangeng, “The role of the 30% threshold for Islamic parties: A fast-growing middle class and religion-based political preferences in Indonesia. Heliyon, 10 (4), e25700,” 2024. doi: 10.1016/j.heliyon.2024.e25700.

Ramjasmaurya, “Social Media Influencer in 2022.” [Online]. Available: https://www.kaggle.com/datasets/ramjasmaurya/top-1000-social-media-channels/data

A. S. J. Eardley-Brunt, L. Song, O. A. M. I. Study, O. A. A. A. Study, and C. Vallance, “Optimising the choice of normalisation method for use in machine-learning classification of human blood plasma ambient ionisation mass spectra,” Int. J. Mass Spectrom., vol. 520, p. 117553, 2026, doi: 10.1016/j.ijms.2025.117553.

Z. Sun and T. He, “Algorithmic Visibility and Informational Quality in Counselling-Related Videos: A Cross-Platform Analysis of YouTube and TikTok,” Computers in Human Behavior Reports, p. 101160, 2026, doi: 10.1016/j.chbr.2026.101160.

A. S. Bafti, M. Vafakhah, V. Moosavi, and H. K. Farsani, “A New Framework for Input Variable Selection Based on the Gamma Test Machine Learning Performance in Quantile Prediction of Flow Duration Curves,” Machine Learning with Applications, p. 100839, 2026, doi: 10.1016/j.mlwa.2026.100839.

C. Ma, B. W. Wang, L. Dai, X. Guan, and Z. Yang, “What determines the effectiveness of social media influencer marketing? An fsQCA-based study of influencer characteristics and content features’ configurational effects,” J. Bus. Res., vol. 189, p. 115170, 2025, doi: 10.1016/j.jbusres.2024.115170.

A. Wiangkham and R. Vongvit, “A hybrid XGBoost and SHAP framework for prioritization and interaction analysis of factors driving metaverse adoption in an engineering context,” Inf. Softw. Technol., p. 108097, 2026, doi: 10.1016/j.infsof.2026.108097.

M. J. Shayegan and N. K. Shiroodi, “A micro-influencer recommender method based on brand features using deep learning approaches,” Telematics and Informatics Reports, vol. 19, p. 100222, 2025, doi: 10.1016/j.teler.2025.100222.

D. Belanche, S. Barta, P. Jordán, and A. Fernández, “TikTok influencers and followers’ engagement: the role of authentic leadership, immersion and inspiration,” Technol. Soc., vol. 87, p. 103414, 2026, doi: 10.1016/j.techsoc.2026.103414.

S. Myers, H. A. Syrdal, R. V Mahto, and S. S. Sen, “Social religion: A cross-platform examination of the impact of religious influencer message cues on engagement–The Christian context,” Technol. Forecast. Soc. Change, vol. 191, p. 122442, 2023, doi: 10.1016/j.techfore.2023.122442.

J.-K. Hsieh, “The impact of influencers’ multi-SNS use on followers’ behavioral intentions: An integration of cue consistency theory and social identity theory,” Journal of Retailing and Consumer Services, vol. 74, p. 103397, 2023, doi: 10.1016/j.jretconser.2023.103397.

P. Eslami, M. Najafabadi, and A. Gharehgozli, “Exploring the journey of influencers in shaping social media engagement success,” Online Soc. Netw. Media, vol. 41, p. 100277, 2024, doi: 10.1016/j.osnem.2024.100277.

S. Qian, Y. Lu, Y. Peng, C. C. Shen, and H. Xu, “Convergence or divergence? A cross-platform analysis of climate change visual content categories, features, and social media engagement on Twitter and Instagram,” Public Relat. Rev., vol. 50, no. 2, p. 102454, 2024, doi: 10.1016/j.pubrev.2024.102454.

K. Spörl-Wang, F. Krause, and S. Henkel, “Predictors of social media influencer marketing effectiveness: A comprehensive literature review and meta-analysis,” J. Bus. Res., vol. 186, p. 114991, 2025, doi: 10.1016/j.jbusres.2024.114991.