Overview
Wikidata is an open knowledge base hosted by the Wikimedia Foundation that can be read and edited by both humans and machines. Serving as the central repository for structured data used by Wikipedia, Wiktionary, Wikisource, and many other projects. Wikidata has become an indispensable resource in both academic research and industrial applications.
In recent years, the growing body of scholarly publications and practical innovations surrounding Wikidata has underscored its evolving role as the backbone of open structured data. Previous workshops have addressed challenges such as data quality, multilingual contributions, community dynamics, and the evolution of collaborative knowledge graphs. However, in recent years, a new branch of research intersecting Wikidata and GenAI has taken place.
The Wikidata Workshop 2025 expands the last years focus to include emerging trends in artificial intelligence and the transformative impact of Generative AI. This year's workshop welcomes submissions that explore the intersection between Wikidata and LLMs, whether in the context of knowledge base construction/completion, data curation, reasoning, or other novel applications.
This workshop brings together everyone working around Wikidata in both the scientific field and industry to discuss trends and topics around this collaborative knowledge graph.
Call for Papers
Topics
We invite researchers from all domains to show the importance of Wikidata in their fields. Topics of interest include, but are not limited to:
- LLM-Enhanced Knowledge Graph Completion: Investigating how LLMs can be used to automatically extract, refine, and augment Wikidata’s vast collection of facts.
- SPARQL Generation Using LLMs to translate natural language questions into Wikidata-compatible SPARQL queries.
- Automated Fact-Checking and Bias Reduction: Utilizing LLMs for real-time verification of Wikidata statements, addressing issues of misinformation, bias, and consistency within the knowledge graph.
- Multilingual and Cross-Lingual Applications: Exploring the potential of LLMs to lower language barriers by enhancing multilingual data curation, translation, and semantic alignment across diverse linguistic communities.
- Human-AI Collaboration in Knowledge Curation: Designing systems where human editors and intelligent bots work together seamlessly, leveraging LLMs to provide editing suggestions, automated validations, and intelligent data enrichment.
- Ethical and Societal Implications: Discussing the challenges and best practices in merging open knowledge graphs with generative AI, including questions of transparency, data provenance, and the societal impact of AI-driven content generation.
- Data Quality and Validation: Detecting inconsistencies, errors, and biases in Wikidata. As well as approaches to analyze entity similarity.
- Multilingual Knowledge Representation: Improving cross-lingual consistency and translation methods.
- AI and Machine Learning Applications: Using Wikidata to train and validate AI models.
- Community and Governance: Analyzing collaboration, contributor behavior, and data policies.
- Wikidata in the Semantic Web: Enhancing interoperability with linked data and external knowledge graphs.
- Automation and Bots: Developing AI-driven agents for data curation and maintenance.
- Wikidata for Science: Supporting open research, citation networks, and metadata management.
Tracks
This workshop will have two tracks: Novel Work, and Previously Published Work.
Papers in the Novel Work track will be published as part of the workshop proceedings. The Previously Published Work track is for papers already published in other conferences, giving the community the chance to access and discuss relevant work that has been presented elsewhere as part of the workshop.
Novel Work Track
The papers will be peer-reviewed by at least three researchers. Selected papers will be published on CEUR (unless authors wish to opt out).
For the Novel Work track, we will accept papers up to 12 pages (excluding references, contribution of the paper should justify the length of the paper). We invite the following types of papers:
Previously Published Work Track
Published papers will be reviewed by the organising committee in terms of topical fit and prominence of the publication venue. They will not be published as part of the proceedings.
For the Previously Published Work track, we will accept papers with no page limit, prioritizing instead the importance and relevance of the publication. We invite the following types of papers:
Submission
TBA (link coming soon).
We ask authors to declare the track they intend on submitting to. To do so, please add, at the beginning of the "title" field on the submission, either the string "[Novel]", for the Novel Work track, or the string "[Published]", for the Previously Published track.
Submission Link: TBA
Important Dates (all deadlines are 23:59 AoE)
Papers submissions: Friday, 11 July 2025
Notification of accepted papers: Thursday, 28 August 2025
Camera ready papers due: Thursday, 07 September 2025
Workshop date: 07 November 2025 in Nara, Japan
Submission Guidelines
Submissions must be as PDF, for the [Novel] track formatted in the style of the CEUR Publications format for CEUR workshop proceedings. A template is available at https://www.overleaf.com/read/pwspggxsbdvy. For the [Published] track, no reformatting of the original PDFs is needed.
Schedule Detail
The workshop time is afternoon: Tentatively 02:00pm - 08:00pm (JST)
All times below in JST
Sessions / Papers
Our Speakers
Daniel Garijo
Technical University of Madrid
Keynote
TBA
Location
Co-located with ISWC 2025
In Nara, Japan, in-person
Organizing Committee
Joint email: 5th-wikidata-workshop@googlegroups.com
Preliminary Program Committee
Elton Figueiredo de Souza Soares, IBM Research, Brazil
Fajar Juang Ekaputra, WU Vienna, Austria
Felipe Vargas Rojas, French National Research Institute for Sustainable Development Montpellier, France
Guilherme Augusto Ferreira Lima, IBM Research, Brazil
Hiba Arnaout, TU Darmstadt, Germany
Jairo Francisco de Souza, UFJF, Brazil
Kholoud Saad Alghamdi, King's College London, UK
Leonardo Guerreiro Azevedo, IBM Research, Brazil
Lucie-Aimée Kaffee, Hugging Face, USA
Lydia Pintscher, Wikimedia Deutschland
Marcelo Tibau de Vasconcellos Dias, UNIRIO, Brazil
Paulo Gimenez, UNIRIO, Brazil
Renato Cerqueira, PUC-Rio, Brazil
Sabrina Kirrane, WU Vienna, Austria
Sandro Rama Fiorini, IBM Research, Brazil
Thomas Pellissier Tanon, Lexistems
Viviane Torres da Silva, IBM Research, Brazil