MUDIDI Framework Uses Language Models to Digitize Multilingual Dictionaries for Endangered Languages
Researchers have introduced MUDIDI, a two-stage AI framework designed to convert scanned multilingual dictionaries into machine-readable formats, with a focus on low-resource and endangered languages. The system benchmarks OCR tools, large language models (LLMs), and vision-language models (VLMs) across 30 public-domain dictionaries spanning diverse writing systems and language families. The work addresses a longstanding barrier to preserving linguistic heritage, as many such dictionaries have remained inaccessible in digital form for decades.
A team of researchers has released MUDIDI, a two-stage framework that leverages modern language models to digitize multilingual dictionaries that exist only as physical scans. Stage One of the framework evaluates character recognition quality and markup preservation, while Stage Two handles dictionary entry segmentation and conversion into SIL's Multi-Dictionary Formatter, a standardized lexicographic schema. The researchers benchmarked traditional OCR systems alongside general-purpose LLMs and vision-language models on a newly released dataset of human-annotated entries drawn from 30 public-domain dictionaries. Results show that LLMs outperform other approaches across most writing systems and languages in both stages of the pipeline. The team also found that providing models with supplementary context—such as a dictionary's introduction—further improves digitization quality. The work is submitted to EMNLP 2026 and the dataset and code are publicly available on GitHub.
What's missing
The preprint does not report quantitative benchmark scores or error rates in the abstract, making it difficult to assess the magnitude of LLM performance gains over OCR baselines. It is also unclear how the framework performs on the most severely endangered or script-unique languages with minimal training data, and whether the approach generalizes to handwritten or degraded scans. The study has not yet undergone peer review.
What different sources said
- arXiv cs.CLCenter
MUDIDI: A Two-Stage Framework for Multilingual Dictionary Digitization with Language Models
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.