Real-Time Agenda Forecasting: A Machine Learning Framework for Predicting Political Agenda Shifts

While Khan et al. (2025) used ML to analyze past opinions, this predicts future agendas. Integrating Guo et al.'s (2023) cross-cutting effects, Boon et al.'s (2023) sentiment analysis, and Gilardi et al.'s (2021) VAR models, we’d train an LSTM network on multi-source data (news, social media, parliamentary records). For example, could it flag a surge in immigration discussions before parliamentary debates? This moves beyond descriptive linkage (Castro Herrero et al., 2021) to anticipatory agenda-setting. The innovation is a dashboard for policymakers/journalists to visualize agenda trajectories, testing whether "agenda prediction" alters media behavior (e.g., preemptive coverage). It reframes agenda-setting as a forecastable phenomenon, not just a retrospective effect.

References:

1. Agenda Setting, Cross-cutting Effects, and Political Expression on Social Media: The Gun Violence Case. Lei Guo, Yiyan Zhang, Kate K. Mays, Afra Feyza Akyürek, Derry Tanti Wijaya, Margrit Betke (2023). Communication Research.
2. First-order linkage analysis (Frequently Applied Designs). Laia Castro Herrero, Theresa Gessler, Sílvia Majó-Vázquez (2021). DOCA - Database of Variables for Content Analysis.
3. Social Media and Political Agenda Setting. F. Gilardi, Theresa Gessler, M. Kubli, Stefan Müller (2021). Political Communication.
4. Hybrid model of machine and deep learning to analyze Twitter data and prediction of online public opinion: revisiting agenda-setting implications. Shumaila Khan, S. Raza, Mubashir Ilyas, Amjad Ali Shah, Umer Zaman, Emenyeonu, C. Ogadimma, Sadaf Sattar (2025). Information Discovery and Delivery.
5. Agencies on the parliamentary radar: Exploring the relations between media attention and parliamentary attention for public agencies using machine learning methods. Jan Boon, J. Wynen, Walter Daelemans, Jens Lemmens, K. Verhoest (2023). Public Administration.