Up to $50,000 Funding
Grant support for language data collection for low-resource European languages.
LINGUA Open CallMicrosoft AI for Good
Building Ukrainian Construction Language for AI
BUDOVA: Building Ukrainian Domain-Specific, Open Voice & Text Archives — the first open Ukrainian construction-domain language dataset for speech and text to advance AI technologies in Ukraine.
01
70+
hours of speech (target)
02
70M+
text tokens (target)
03
367B
USD reconstruction
04
CC-BY 4.0
license
BUDOVA Pipeline01 / 04
input
""
tokens
арматурнийкаркасфундаменту
entities
MATарматурний
ELMкаркас
ELMфундаменту
confidence
—
Initiative
LINGUA Open Call
A Microsoft AI for Good Lab initiative under EU Digital Unlock, supporting digital inclusion for low-resource European languages and building language resources for 10 European languages.
Microsoft
Azure Compute Resources
Cloud computing credits for up to 2 years for data processing and validation.
Technical Support
Research collaboration with AI for Good Lab, EPFL, and ETH Zürich.
Open Models
Integration with Apertus, EuroLLM, SmolLM3 and other multilingual models.
Challenge
Why This Is Critical
Despite 30–46 million speakers, Ukrainian remains critically underrepresented in AI technologies — especially in specialized domains.
<0.6%
of web content in Ukrainian
Digital Exclusion
No technical language datasets, speech recognition systems for construction sites, or AI tools supporting Ukrainian building codes.
3
dialect groups, 15+ subdialects
Dialect Diversity
Northern, Southwestern (Volynian, Galician, Podilian, Transcarpathian, Bukovynian, Hutsul, Boiko, Lemko), and Southeastern dialect groups remain undocumented in technical contexts — AI systems risk failure for non-standard speakers.
15–25%
performance gap
Morphological Complexity
Cyrillic script, 7 cases, 3 genders, moveable stress — unique tokenization challenges with a performance gap vs. English.
What We Build
Open Resources for Construction
The first comprehensive Ukrainian technical language dataset with dialectal coverage across all major regions of Ukraine.
Speech
Speech Dataset
100+ hours of annotated construction speech with transcriptions, covering Northern, Southwestern, and Southeastern Ukrainian dialects.
<25%WER
>0.70Cohen's κ
Text
Text Corpus
100M+ tokens of construction documentation — DBN codes, technical specifications, safety protocols with NER annotations.
>0.85NER F1
5–10Knew terms
Infrastructure
Technical Infrastructure
Baseline ASR and NER models, data processing pipelines, documentation per Datasheets for Datasets standards. Available on Hugging Face & Zenodo.
WAV/FLACLossless
DOIPersistent
Projected Impact
Roadmap
24 Months, 5 Phases
Phased deployment with quarterly releases for community feedback.
01 / 05Phase 1
Infrastructure & Preparation
Months 1–3 · Jan – Mar 2026Infrastructure setup, partnership agreements, GDPR protocols, participant recruitment.
02 / 05Phase 2
Data Collection
Months 4–15 · Apr 2026 – Mar 2027Phased collection: 25h speech + 25M tokens quarterly until reaching 100+h and 100M+ tokens.
03 / 05Phase 3
Annotation & Validation
Months 6–18 · Parallel with collectionTwo-stage validation: crowdsourced transcription, then expert review. Inter-annotator agreement >0.70.
04 / 05Phase 4
Baseline Model Development
Months 12–21 · Jan – Sep 2027Training ASR and NER models on collected data. Achieving WER <25% for construction terminology.
05 / 05Phase 5
Release & Transfer
Months 21–24 · Sep – Dec 2027Final dataset release on Hugging Face and Zenodo with DOIs. Documentation, sustainability transition.
Datasets
Open Research Resources
Large-scale datasets and benchmarks for training, evaluating, and testing Ukrainian construction-domain NLP models.
Speech Dataset
100+hours of speech
Annotated construction speech recordings with transcriptions covering all major Ukrainian dialects — on-site discussions, briefings, and consultations from across Ukraine.
Hugging FaceText Corpus
100M+text tokens
Construction documentation with NER annotations — DBN codes, technical specifications, safety protocols, and project documentation.
Hugging FaceConstruction Lexicon
5–10Kdomain terms
Domain-specific terminology covering building materials, techniques, safety standards, and regulatory language across construction sub-domains.
Hugging FaceLanguageUkrainian (uk-UA) · Northern, Southwestern & Southeastern dialects
LicenseCC-BY 4.0 (data) · Apache 2.0 (models) · MIT (code)
HostingHugging Face · Zenodo (DOI) · GitHub
PrivacyGDPR · Voice anonymization · PII removal
Examples
Live examples from the corpus.
How BUDOVA annotates Ukrainian construction language — entity, register, and context labels (illustrative examples; the fine-tuned model ships with v1.0).
«Monolithic reinforced-concrete load-bearing structures of elevated responsibility class must be designed accounting for seismic loads per DBN V.1.1-12:2014.»
- Register
- Formal · normative
- Material
- Monolithic reinforced concrete
- Source
- DBN V.1.1-12:2014
- NER tags
- 4 entities · 0 ambiguities
Open Resources
Full Transparency. No Restrictions.
All outputs — data, models, code — under open licenses.
Datasets
Speech and text corpora with full annotations, metadata, and a construction lexicon of 5–10K terms.
Baseline Models
Fine-tuned ASR model (Whisper / Wav2Vec2) and NER models for the construction domain with reproducible scripts.
Infrastructure
Processing pipelines, anonymization utilities, validation workflows — for replication by other low-resource projects.
Ready to Get Started?
Download our datasets from Hugging Face and start building with Ukrainian construction language data.
Browse All DatasetsHow to Cite
If you use BUDOVA resources in your research, please cite:
@misc{budova2026,
title = {BUDOVA: Building Ukrainian Domain-Specific, Open Voice & Text Archives},
author = {Dolhopolov, Serhii},
year = {2026},
publisher = {Hugging Face & Zenodo},
license = {CC-BY-4.0},
url = {https://huggingface.co/datasets/budova}
}Deliverables
Project Outputs
Concrete research outputs and open resources produced throughout the project lifecycle.
01 / 04
3datasets
Dataset Releases
Versioned speech and text datasets published on Hugging Face and Zenodo with full documentation and DOIs.
25%ETA Q4 2026
- Infrastructure & GDPR protocols
- First 25h speech collected
- Text corpus v0.1
02 / 04
2+models
Baseline Models
Fine-tuned ASR (Whisper / Wav2Vec2) and NER models for Ukrainian construction domain with reproducible scripts.
5%ETA Q2 2027
- Model architecture selected
- Training pipeline setup
- Benchmark evaluation
03 / 04
3+papers
Scientific Papers
Peer-reviewed publications at NLP and AI conferences documenting methodology, benchmarks, and findings.
10%ETA Q3 2027
- Literature review completed
- Methodology paper draft
- Conference submission
04 / 04
100%open-source
Open-Source Tools
Processing pipelines, anonymization utilities, validation workflows, and documentation per Datasheets for Datasets standards.
15%ETA Q4 2027
- Repository structure created
- Anonymization pipeline
- Validation workflows
Research Group
Project Team
An interdisciplinary team with expertise in AI, NLP, construction, cybersecurity, and data management.
01 / 01Project Team Lead
Serhii Dolhopolov
KNUCA AI Lab · Ph.D. Student in Computer Science
AI researcher and entrepreneur specializing in Natural Language Processing. Founder of KernelGlide — AI solutions for 10+ clients in construction. Principal Investigator on state-funded research (UAH 2.9M) in multimodal content analysis. Author of the textbook "Modeling AI Tasks" (546 pp.).
Top Contributors
- 0100annot.Be the first contributor!/annotate
Host institution
knuba.edu.uaKyiv National University of Construction and Architecture
Founded in 1930 · 95 years of leadership in construction education & research