Bibliographic record
Abstract
Purpose Customers often communicate with a company by sending it emails, filling out the “contact us” form on the company’s website, and posting messages on social media platforms. Unfortunately, a substantial part of these incoming customer messages can be spam. Whether an email is authentic (i.e. ham) or spam is specific to the company. Consequently, commercial spam classifiers (e.g. the one Microsoft Outlook uses) do not work well. A machine learning-based spam classifier can solve this problem. Design/methodology/approach I used a large dataset from a publicly traded retailer in the United States to test 27 TF-IDF-based and six word embedding-based binary classifiers. Findings I found that RoBERTa – a sophisticated embedding-based classifier – provided the lowest false positive rate of 5.31%. However, its rival classifier – XLNet – offered marginally superior (92%) balanced accuracy, relative to RoBERTa’s 91%. To enhance the use of the models by other organizations and academics, I offer supplemental models trained on only external datasets and a combination of internal and external datasets. Research limitations/implications The manuscript contributes by applying a broad set of existing machine learning models to solve a real problem for a real company. Practical implications I publish my code (via the journal), which other companies and researchers can use. Originality/value The manuscript is original in comparing a broad range of machine learning classifiers on real data from a real company.
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 | 0.003 | 0.014 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
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".