HomoEnsNER: Does Language Alignment Outperform Architectural Complexity in Gujarati Named Entity Recognition?
Summary
This paper proposes HomoEnsNER, a homogeneous ensemble of five GujaratiBERT models for Gujarati named entity recognition, and shows it outperforms heterogeneous alternatives that rely on architectural diversity, achieving state-of-the-art F1 on the Naamapadam test split.
View Cached Full Text
Cached at: 08/05/26, 07:44 AM
# HomoEnsNER: Does Language Alignment Outperform Architectural Complexity in Gujarati Named Entity Recognition? Source: [https://arxiv.org/abs/2608.03105](https://arxiv.org/abs/2608.03105) [View PDF](https://arxiv.org/pdf/2608.03105) > Abstract:Named Entity Recognition \(NER\) for Gujarati remains underexplored, hindered by the absence of capitalization cues, rich morphology, lexical ambiguity, and free word order\. Prior ensemble work has emphasized architectural diversity by combining heterogeneous classifiers, multilingual encoders, or classical sequence models, rather than exploiting language\-aligned monolingual pretraining\. This study asks whether, for a low\-resource, morphologically rich language like Gujarati, a homogeneous ensemble of a single monolingual encoder outperforms such architectural diversity\. We propose HomoEnsNER, a homogeneous ensemble of five independently fine\-tuned GujaratiBERT models combined via majority voting, evaluated against a single GujaratiBERT baseline and six heterogeneous alternatives, including combinations with MuRIL\-base, MuRIL\-large, IndicBERT, mBERT, BiLSTM, CRF, and a stacked BiLSTM\-CRF\-GujaratiBERT architecture\. All eight models were trained under a consistent budget and evaluated using entity\-level F1 on the Naamapadam Gujarati test split\. HomoEnsNER achieved the highest F1 \(0\.8442\), surpassing the baseline \(0\.8347\) and every heterogeneous alternative \(lowest: 0\.7855\), indicating that language alignment is a more effective, budget\-conscious ensembling strategy than architectural complexity for low\-resource Indian language NER\. ## Submission history From: Chandrakant Bhogayata \[[view email](https://arxiv.org/show-email/174cfdb6/2608.03105)\] **\[v1\]**Tue, 4 Aug 2026 04:21:35 UTC \(351 KB\)
Similar Articles
BERT-based Models vs. Large Language Models for Low-Resource Named Entity Recognition: A Comparative Study on Marathi
This paper compares fine-tuned MahaBERT-based models with large language models (Gemini, LLaMA-3.3-70B, Gemma) for Marathi named entity recognition, finding that the specialized BERT models significantly outperform the LLMs, achieving F1-scores of 0.88–0.91 versus 0.57–0.69.
Language Specificity vs. Domain Diversity: Benchmarking Transformers for Bangla Medical NER
This paper benchmarks transformer models for Bangla medical Named Entity Recognition, demonstrating that fine-tuned XLM-RoBERTa achieves state-of-the-art performance and outperforms language-specific models, emphasizing the importance of domain diversity in low-resource settings.
Enhancing Biomedical Named Entity Recognition via Multiple Programming Languages Instruction Tuning and Ensemble Method
MITE reformulates biomedical NER as a structure-to-structure generation task by representing instructions and entity outputs in multiple programming-language formats (Python, C++, Java), using entity-level voting at inference to improve robustness. Experiments on six BioNER datasets show it outperforms BERT- and LLM-based baselines with strong cross-dataset generalization.
Structure-Guided Entity Resolution: Fine-Tuning LLMs for Robust Name Matching in Complex Linguistic Contexts
This paper presents Structure-Guided Entity Resolution (SGER), a framework that fine-tunes LLMs through curriculum learning for robust person name matching in linguistically diverse contexts, achieving 99.02% accuracy on Indian identity data and deployed at Dream11.
NE-BERT: A Multilingual Language Model for Nine Northeast Indian Languages
NE-BERT is a multilingual encoder model trained on 8.3 million sentences for nine Northeast Indian languages and two anchor languages, outperforming existing models like IndicBERT-V2 and MuRIL with significant gains on ultra-low-resource languages through custom tokenization.