Fusion of Fast-text and Indo-Wordnet for Disambiguation of Word Sense in the Marathi Language
Bibliographic record
Abstract
This research employs the combination of the FastText model and Indo-WordNet to address the issue of word sense disambiguation (WSD) in Marathi literature. The initial iteration of the algorithm employed word pair matching as the technique to ascertain the presence of overlap between the items in the "context bag" and the "sense bag" derived from the lexical resource WordNet. The current methodology involves the computation of overlap by utilizing a semantic similarity metric that leverages fastText subword embeddings. This approach demonstrates proficiency in effectively managing unanticipated word formations, while simultaneously elucidating the inherent semantics of the terms. Significant progress has been achieved in the field of Word Sense Disambiguation (WSD) for both the English language and many European languages. There is a substantial challenge to be surmounted in relation to Marathi and other languages spoken in India. The Marathi text corpus, sourced from the government of India, comprises a vast assemblage of Marathi sentences. The dataset used in this study consisted of the Indo WordNet for the Marathi language and the Marathi Online Dictionary. The results of the conducted experiments demonstrate promising discoveries. The target words that possess semantically distinct synsets in WordNet are assigned a high F1 score. The achieved F1 score of 89% above the baseline and signifies substantial advancements in compared to previous knowledge-based methodologies employed for low resource Indian languages.
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 | 0.005 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.009 | 0.005 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.009 |
| Open science | 0.001 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.004 |
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".