GraphAush: Combining Adversarial Learning and Graph Embedding for a Novel Shilling Attack Model Towards Robust Recommender Systems
Bibliographic record
Abstract
Recommender systems (RS) are integral to modern e-commerce and content platforms. Yet, their reliance on user-item interaction data makes them vulnerable to shilling attacks, where fake data is injected to manipulate recommendations. Traditional shilling attack strategies utilize basic statistical properties of user-item data to create deceptive profiles. Still, recent advancements have shifted towards model-based attacks leveraging machine learning to enhance effectiveness and evade detection. This paper introduces GraphAush, a novel neural shilling attack model that employs Generative Adversarial Networks (GANs) with a novel generator shilling loss tailored to manipulate the user-item interaction graph. By incorporating a novel generator shilling loss function that leverages Node2Vec embeddings, GraphAush optimizes fake profile generation to maximize the effectiveness of attacks while minimizing detectability. This method overcomes the limitations of previous models, which either require intricate knowledge of the target RS or use indirect graph-based approaches. The efficacy of GraphAush was validated through experiments with multiple benchmark datasets, revealing its strong performance against traditional heuristic and other GAN-based attack methods. This innovative approach highlights a significant advancement in adversarial techniques for RS and sets a new benchmark for evaluating shilling attack strategies.
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 | 0.002 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".