Twitter Bot Detection Using Neural Networks and Linguistic Embeddings
Bibliographic record
Abstract
Twitter is a web application playing the dual role of online social networking and micro-blogging. The popularity and open structure of Twitter have attracted a large number of automated programs, known as bots. In this article, we propose a Twitter bot detection model using recurrent neural networks, specifically bidirectional lightweight gated recurrent unit (BiLGRU), and linguistic embeddings. To the best of our knowledge, our Twitter bot detection model is the first that does not require any handcrafted features, or prior knowledge or assumptions about account profiles, friendship networks or historical behavior. The proposed model uses only textual content of tweets and linguistic embeddings to classify bot and human accounts on Twitter. Experimental results show that the proposed model performs better or comparably to state-of-the-art Twitter bot detection models while requiring no feature engineering, making it faster and easier to train and deploy in a real network. We also present experimental results that show the performance and computational costs of different types of linguistic embeddings and recurrence network variants for the task of Twitter bot detection. The results will potentially help researchers design high-performance deep-learning models for similar tasks.
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 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".