Ensemble Learning Approaches for SMS Spam Detection: A Comparative Study of Text Classification Models
Bibliographic record
Abstract
For users who rely on single-use mobile phones, the global problem of receiving unwanted marketing messages through SMS remains a significant concern. In recent years, extensive use of machine learning and deep learning approaches has been explored to address this challenge. To improve predictive accuracy, the outputs of multiple models were combined using a majority-voting strategy. This work presents a comparative analysis of several text classification techniques, highlighting the importance of reliably identifying and labeling spam SMS messages. After data preprocessing, messages were transformed into numerical representations using TF-IDF, which emphasizes uncommon but informative terms over frequent ones. Among the tested methods, the Relevance Vector Machine achieved the strongest performance in the data, reaching an F1 is 0.975176. In addition, this examined alternative spam detection algorithms, including Logistic Regression, XGBoost, and LightGBM. The preprocessing pipeline included duplicate removal, text normalization with spaCy, label encoding, and TF-IDF vectorization. Two experimental conditions were evaluated: one without handling class imbalance and another with imbalance adjustment. Results showed that ensemble-based methods, particularly Gradient Boosting, XGBoost, and LightGBM, consistently delivered superior performance. Under imbalanced data conditions, both XGBoost and LightGBM achieved F1 scores of 0.99 across the majority and minority classes. When class imbalance was corrected, their performance remained strong, with F1 scores of 0.98 for all classes. Logistic Regression also demonstrated robust results, confirming its role as a reliable baseline. Overall, the findings indicate that the proposed RVM framework is effective for SMS spam classification and has practical applicability in real-world scenarios.
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 | 0.006 | 0.008 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.004 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| 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".