Fine-Tuned IndoBERT Based Model and Data Augmentation for Indonesian Language Paraphrase Identification
Bibliographic record
Abstract
Natural Language Processing tasks in the Indonesian language have recently flourished thanks to the research of IndoBERT and its benchmark.Despite being the fourth most used language over the internet, the Indonesian language NLP task still has some gaps, one of them being the Paraphrase Identification task.In order to solve this gap, we proposed a fine-tuned IndoBERT based model for Paraphrase Identification.Several methods have been researched in this paper from setting the baseline, Data Augmentation, fine-tune the classifier, and task reformulation.Besides the model, this paper also provides the Paraphrase Identification dataset in Indonesian language.The baseline IndoBERT model performs well, it proves that IndoBERT is one of the fittest methods to use.We then researched further and proposed a Modified Easy Data Augmentation that augments very well in this task and potentially on other NLP tasks.We compared traditional machine learning classifiers with deep neural network classifiers, fine-tuned them, and selected the best classifier for this task.Furthermore, we tried entailment task reformulation.The Modified EDA shows a successful augmentation that increases both accuracy and F1 score for all the models.A slightly complex upgrade for the classifier also increased the performance while maintaining a reasonable training time.
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 | 0.001 | 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 | 0.001 |
| 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".