RanABD: MTD-Based Technique for Detection of Advanced Session-Replay Web Bots
Bibliographic record
Abstract
In the current digital landscape, cyberattacks have become increasingly sophisticated in their attempts to evade detection. One such example is the session-replay web bot attack, where hackers use previously recorded human mouse movements (i.e., sessions) to emulate human behavior on the target web sites and apps. With the emergence of advanced AI, hackers are further expected to utilize these programs to generate carefully randomized session-replay bots that still exhibit human-like behavior but without replaying/repeating identical mouse trajectories, as was previously the case. Detecting such advanced bots in the traditionally designed web pages and sites is exceptionally hard if not impossible. In this paper, we propose RanABD, a novel defensive web page randomization technique that is built on the concepts of moving-target defence (MTD) and is designed to counter advanced session-replay web bots. RanABD performs randomized micro modifications in the alignment of select visual HTML elements and element attributes in the target web page, while causing minimal disturbances in the page's overall appearance and functionality. By doing so, the technique ensures that the distances between trajectories of genuine human-visitors, as well as trajectories of repeat visits by the same human user, are sufficiently separated in the Feature Space. For session-replay bot operators, the only way to bypass this defence is by increasing the degree of randomization in replay sessions, but this approach is likely to backfire as it inevitably results in outlier-like trajectories that are even easier to detect. According to our knowledge, this is the first research paper that explicitly addresses the issue of advanced session-replay bots as well as proposes a novel technique that can effectively detect these specific types of bots.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".