Author(s)

ZhiYang Chen^1*, Wei Ding², GengYan Li²

Affiliation(s)

¹School of Mathematics and Physics, Southwest University of Science and Technology, Mianyang 621000, Sichuan, China.

²School of Computer Science and Technology, Southwest University of Science and Technology, Mianyang 621000, Sichuan, China.

Corresponding Author

ZhiYang Chen

ABSTRACT

Addressing the highly discrete and complex patterns of user behavior on social platforms, this study constructs a predictive framework combining ensemble learning with dynamic time series decomposition to achieve precise modeling of user online status and interaction behaviors. Within the user online status recognition and multi-level influencer recommendation models, the study first extracts key features such as the user's past 3-day, 7-day, and historical total active days. Utilizing a random forest classifier to capture the nonlinear correlations in user behavior, it effectively determines the online status for specific dates. Subsequently, by defining an interaction score formula incorporating time-decay weights and integrating user preferences with global popularity, the model efficiently recommends interactive bloggers, achieving a recall rate of 80.48% on the validation set. In precise temporal interaction prediction based on user segmentation and time series decomposition, this study addresses the temporal periodicity of historical behavior by introducing the Facebook Prophet model. Through automatic decomposition of trend and seasonal components, it captures user activity patterns during holidays and specific periods. To enhance prediction accuracy, the study further clusters users using the K-means algorithm and configures model hyperparameters based on behavioral feature differences. Experimental results demonstrate that this approach effectively addresses the challenge of predicting interaction intensity across time intervals under sparse data conditions, providing scientific decision support for personalized platform operations.

KEYWORDS

Random forest; Facebook prophet; User behavior modeling

CITE THIS PAPER

ZhiYang Chen, Wei Ding, GengYan Li. Predicting user activity and interaction sequences on social platforms by integrating ensemble learning and time series decomposition. Trends in Social Sciences and Humanities Research. 2026, 4(1): 36-41. DOI: https://doi.org/10.61784/tsshr3208.

REFERENCES

[1] Hartl T, Hutter C, Weber E. Matching for three: The search activities of workers, firms, and employment services. Economic Modelling, 2026, 155: 107434. DOI: 10.1016/J.ECONMOD.2025.107434.

[2] Vineela K, Grover J. A Comprehensive Evaluation of Health Hazard Prevention Education Programs Targeting Online Activity Risks Among Students. Journal of Educational Research and Policies, 2025, 7(11): 3-9. DOI: 10.53469/JERP.2025.07(11).02.

[3] Schutt K R, O’Donoghue L A, Hargraves L J, et al. Diminished mental well-being in the COVID-19 pandemic: The role of infection risk, social disconnection, and income loss at the individual and neighborhood levels. Wellbeing, Space and Society, 2025, 9: 100302. DOI:10.1016/J.WSS.2025.100302.

[4] Ron D A J, Whitehead J, Kaufmann J, et al. Gender and Sexual Orientation Differences in the Relationship Between Social Media Use and Disordered Eating: Results From a Serial Cross-Sectional Youth Survey From 2022 to 2024. The Journal of adolescent health: official publication of the Society for Adolescent Medicine, 2025, 78(1): 110-118. DOI: 10.1016/J.JADOHEALTH.2025.08.022.

[5] Lu Y, Quan K. Effects of doctors’ participation in internet live streaming free medical consultations on their benefits: empirical evidence from a quasi-natural experiment. Journal of Business Research, 2025, 199: 115560. DOI: 10.1016/J.JBUSRES.2025.115560.

[6] Soltani M, Dara R, Poljak Z, et al. Leveraging social media and google trends to identify waves of avian influenza outbreaks in USA and Canada. Expert Systems With Applications, 2025, 291: 128482.

[7] Ghermandi A, Depietri Y, Orenstein E D. Human-nature interactions through a digital prism: Heterogeneity in user socio-demographics and content across social media platforms. Landscape and Urban Planning, 2026, 268: 105568.

[8] Aydin T, Parris A B, Arabaci G, et al. On the relationship between internet addiction and ADHD symptoms in adults: does the type of online activity matter?. BMC public health, 2025, 25(1): 2072. DOI: 10.1186/S12889-025-23040-4.

[9] Slote K, Daley K, Succar R, et al. How advocacy groups on Twitter and media coverage can drive US firearm acquisition: A causal study. PNAS nexus, 2025, 4(6): pgaf195. DOI: 10.1093/PNASNEXUS/PGAF195.

[10] Barrow G. GETTING TRUMPED. Commercial Motor, 2025, 236(6152): 38-39.