Correcting Arabic OCR Errors Using Improved Topic-Based Language Models
Bibliographic record
Abstract
The OCR output of scanned document images suffers from recognition errors especially when dealing with languages that are characterized by particularities and rich morphology such as the Arabic language, thus an effective error correction model is greatly needed. This paper focuses on three aspects of post-processing correction. First, improving the alignment and error n-gram models by adding correction rules based on character meta-classes rather than on specific characters, which is more suitable for the Arabic language. Second, using the language models to understand and correct the Arabic word fragment resulting from agglutinated affixes or isolated letters. The last will concern improving the language models by adding semantic information to the correction process, by using the bidirectional n-grams, stemming and removing stop words, which gives higher weights to n-grams sharing semantic meanings. In addition, we use a topic corpus, not a global one for a better probability distribution. The proposed model is effective in correcting the lexical errors and covered the semantic ones, that were not frequently reported by OCRs and are corrected after a manual proofreading. The proposed method shows an increase in the correction rate of almost 13% especially in meaningful terms.
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 | 0.002 | 0.007 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.002 | 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".