Modeling Control Agents in Social Media Networks Using Reinforcement Learning
Bibliographic record
Abstract
Designing efficient control strategies for opinion dynamics is a challenging task. Understanding how individuals change their opinions in social networks is essential to countering malicious actors and fake news and mitigating their effect on the network. In many applications such as marketing design, product launches, etc., corporations often post curated news or feeds on social media to steer the users’ opinions in a desired way. We call such scenarios opinion shaping or opinion control whereby a few selected users, called control users, post opinionated messages to drive the others’ opinions to reach a given state. In this paper, we are interested in the control of opinion dynamics in social media using a combination of multi-agent systems and Q-learning. The social media environment is modeled with flexible multi-agent opinion dynamics that can capture the interaction between individuals in social media networks using a two-state updating mechanism. The environment is formulated as a partially observable multi-agent Markov decision process. We propose using intelligent reinforcement learning (RL) agents to control and shape the social network’s opinions. We present the social media network as an environment with different kinds of individuals and connections and the influencing agent as an RL agent to influence the network.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".