Experimental Evaluation of Baselines for Forecasting Social Media Timeseries

TL;DR

This paper compares four baseline methods for forecasting social media activity across different platforms and geopolitical contexts, providing insights into their relative accuracy and guiding future social media modeling efforts.

Contribution

It offers an experimental evaluation of multiple baseline forecasting methods on diverse social media datasets, highlighting their strengths and limitations.

Findings

Certain baselines outperform others depending on the metric.

Baseline performance varies across platforms and geopolitical contexts.

Guidance for selecting appropriate baselines in future social media forecasting studies.

Abstract

Forecasting social media activity can be of practical use in many scenarios, from understanding trends, such as which topics are likely to engage more users in the coming week, to identifying unusual behavior, such as coordinated information operations or PumpNDump efforts. To evaluate a new approach to forecasting, it is important to have baselines against which to assess performance gains. We experimentally evaluate the performance of four baselines for forecasting activity in several social media datasets that record discussions related to three different geo-political contexts synchronously taking place on two different platforms, Twitter and YouTube. Experiments are done over hourly time periods. Our evaluation identifies the baselines which are most accurate for particular metrics and thus provide guidance for future work in social media modeling.