WDAqua · Project

AI-Powered Question Answering System That Finds Answers Across Web Data

digitalPrototypeTRL 4Thin data (2/5)

Imagine you could ask a computer a plain question — like you'd ask a colleague — and it would search through millions of web databases, company records, and public datasets to find the exact answer. That's what WDAqua built. Instead of spending hours clicking through websites and spreadsheets, their system understands your question, figures out where the data lives, and brings back a direct answer. Think of it as a smart search engine that actually answers questions instead of just giving you a list of links.

By the numbers

consortium partners

countries represented

industry partners in consortium

SMEs in consortium

38%

industry ratio in consortium

total project deliverables

demo deliverables (OpenQA demonstrator + core platform)

The business problem

What needed solving

Companies drown in data spread across websites, databases, and public records, but finding specific answers still requires manual searching and technical expertise. Employees waste hours navigating complex query interfaces or scrolling through irrelevant search results instead of getting direct answers to straightforward questions.

The solution

What was built

The project built the OpenQA platform — a question answering system that integrates natural language processing, entity recognition, data quality assessment, dataset linking, and query construction into a self-contained demonstrator. It also produced integrated tools for data cleaning, enrichment, cross-domain dataset alignment, and dynamic dataset indexing across 12 deliverables.

Audience

Who needs this

E-commerce platforms needing smarter product search and customer Q&ACity governments building citizen-facing smart city information servicesPublishers and media companies wanting automated content Q&A systemsEnterprise data teams building internal knowledge search toolsCustomer support departments automating FAQ and help desk responses

Business applications

Who can put this to work

E-commerce & Retail

any

Target: Online retailers and marketplace operators

If you are an e-commerce company dealing with customers who struggle to find the right products across your catalog — this project developed the OpenQA platform that understands natural language questions and searches across product databases to deliver precise answers. Instead of customers abandoning their carts because search failed them, the system interprets what they actually mean and connects them to the right product data.

Public Sector & Government

enterprise

Target: Municipal authorities and public service agencies

If you are a city administration dealing with citizens who cannot find information across scattered government databases — this project developed question answering technology tested in smart city settings that lets people ask plain questions and get answers pulled from multiple public data sources. The OpenQA demonstrator integrates entity recognition, data linking, and natural language processing into one self-contained system.

Publishing & Media

mid-size

Target: Digital publishers and content platforms

If you are a publishing company dealing with readers who need fast, accurate answers from your content archives — this project developed data-driven question answering tools that can search across linked datasets, recognize entities, and return direct answers instead of generic search results. The platform includes data quality assessment and dynamic dataset indexing built by a consortium of 13 partners across 5 countries.

Frequently asked

Quick answers

What would it cost to implement this question answering technology?

The project was funded as a Marie Skłodowska-Curie training network (MSCA-ITN), and specific licensing or implementation costs are not published in the available data. You would need to contact the coordinator at the University of Bonn or one of the 5 industry partners to discuss commercial terms.

Can this scale to handle enterprise-level data volumes?

The OpenQA core platform was designed to handle cross-domain datasets with integrated indexing, alignment, and linking tools for dynamic datasets. The system includes quality-driven crawling services and data enrichment solutions. However, enterprise-scale stress testing results are not detailed in the available deliverables.

What is the IP situation — can I license this technology?

WDAqua was an MSCA-ITN project, meaning IP generated is subject to EU grant agreement rules. With 5 industry partners and 3 SMEs in the consortium, some components may have commercial licensing paths. Contact the coordinator or the specific industry partner whose module you need.

How mature is the OpenQA platform — is it production-ready?

The project produced an OpenQA core platform (prefinal) and a complete self-contained demonstrator integrating modules from multiple research teams. Based on available project data, this reached demonstrator stage with integrated NER, disambiguation, and query construction components, but it was primarily a research and training output.

Does this work with existing enterprise databases and systems?

The platform was built to work with linked data and open data on the web, with specific tools for dataset alignment, indexing, and quality assessment. Integration with proprietary enterprise databases would likely require customization. The consortium included 5 industry partners who contributed to real-world application testing.

What languages and data types does the system support?

Based on the project objectives, the system covers natural language processing, speech recognition, entity recognition, and data visualization. The consortium spanned 5 countries (DE, EL, ES, FR, UK), suggesting multilingual capability, though specific language coverage is not detailed in the available data.

Consortium

Who built it

WDAqua brought together 13 partners from 5 countries (Germany, Greece, Spain, France, UK), with a healthy 38% industry ratio — 5 industry partners and 3 SMEs alongside 5 universities and 2 research institutes. This mix means the research was tested against real business needs, not just academic benchmarks. The industry partners contributed to application areas including e-commerce, public sector information, publishing, and smart cities. For a business looking to adopt this technology, the industry partners would be the first point of contact, as they have hands-on experience applying these research outputs to commercial settings.

RHEINISCHE FRIEDRICH-WILHELMS-UNIVERSITAT BONNCoordinator · DE
ATHENS TECHNOLOGY CENTER ANONYMI VIOMICHANIKI EMPORIKI KAI TECHNIKI ETAIREIA EFARMOGON YPSILIS TECHNOLOGIASpartner · EL
ETHNIKO KAI KAPODISTRIAKO PANEPISTIMIO ATHINONparticipant · EL
UNIVERSITE JEAN MONNETparticipant · FR
WOLTERS KLUWER DEUTSCHLAND GMBHpartner · DE
UNIVERSITY OF SOUTHAMPTONparticipant · UK
OPEN DATA INSTITUTE LBGparticipant · UK
FUNDACIO EURECATpartner · ES
GOTTFRIED WILHELM LEIBNIZ UNIVERSITAET HANNOVERparticipant · DE

How to reach the team

Coordinator is Rheinische Friedrich-Wilhelms-Universitat Bonn (Germany). Use Google AI Search to find the project coordinator's direct contact details.

Order a report on this consortium See RHEINISCHE FRIEDRICH-WILHELMS-UNIVERSITAT BONN profile →

Next steps

Talk to the team behind this work.

Want to explore how WDAqua's question answering technology could work for your business? SciTransfer can connect you directly with the research team and help assess fit for your specific use case.

Order a report on this topic More in Digital & ICT