Call for papers

6th International Workshop on Scientific Knowledge: Representation, Discovery, and Assessment (Sci-K 2026)

web: https://sci-k.github.io, X: @scik_workshop, LinkedIn: Sci-k workshop

Held in conjunction with the International Semantic Web Conference (ISWC) 2026, Bari (Italy)

Aim and Scope:

Recently, we have experienced a massive increase in the volume of scientific articles and research artefacts (e.g., datasets, models, software packages). This trend is expected to continue and pose challenges, including developing large-scale machine-readable representations of scientific knowledge, making scholarly data and knowledge discoverable and accessible, and designing reliable and comprehensive metrics to assess scientific impact and measure the quality of structured scientific resources and AI-driven research support. Sci-K provides a forum for researchers and practitioners from diverse disciplines to present, educate, and guide research on scientific knowledge. Three themes cover the most important challenges in this field:

Representation

There is a need for flexible, context-sensitive, fine-grained, and machine-actionable representations of scholarly knowledge that are, at the same time, structured, interlinked, and semantically rich: Scientific Knowledge Graphs (SKGs), also known as Research Knowledge Graphs (RKGs). Even more so, in line with the recent Barcelona Declaration on Open Research Information, SKGs/RKGs can power data-driven services to navigate, analyse, and make sense of research dynamics, thus becoming the structural backbone of model scholarly communication and research intelligence, such as AI-driven research assistants. Current challenges relate to the design of ontologies or alternative representation methods that conceptualise scholarly knowledge, model its representation, both metadata as well as richer semantic content such as hypotheses, methods, claims, and research results, and enable exchange. Furthermore, supporting interdisciplinary knowledge representation and cross-domain alignment across heterogeneous scientific fields remains a key challenge. Lastly, application domains such as semantic publishing illustrate how representation approaches can be operationalised in scholarly communication, while also exposing open challenges related to usability, adoption, and the balance between structured and natural language formats.

Discoverability

Scholarly information should be easily findable, discoverable, and visible so that it can be mined and organised within SKGs/RKGs. Discovery tools should be able to crawl the Web and identify scholarly data, whether on a publisher’s website or in institutional repositories, preprint servers, or open-access repositories. This is challenging and requires a deep understanding of both the scholarly communication landscape and the needs of a range of stakeholders: researchers (across different fields and subfields), publishers, funders, and the general public. Other challenges include the discovery and extraction of entities and concepts, the integration of information from heterogeneous sources, the identification of duplicates, the identification of connections between entities, and the identification of conceptual inconsistencies. We are particularly interested in modern systems that integrate AI, NLP, and LLM technologies, including hybrid human-AI workflows where automated methods are combined with expert curation and validation. Lastly, application domains and use cases are needed to better understand for which concrete research tasks ontologies, knowledge graphs, and LLMs can effectively support researchers, such as literature exploration, hypothesis generation, and synthesis of scientific knowledge.

Assessment

Due to the continuous growth in the volume and diversity of research products, and the global movement around Responsible Research Assessment reforms (e.g., DORA, CoARA), inclusive approaches to research evaluation are more relevant than ever. There is a need for reliable, comprehensive, inclusive and equitable metrics and indicators of the scientific impact and merit of publications, datasets, research institutions, individual researchers, and other relevant entities. In addition, there is a growing need for methods to assess the quality, reliability, and usefulness of the underlying representations and discovery systems themselves, including scientific knowledge graphs, ontologies, and AI-driven discovery tools, in terms of their coverage, accuracy, interpretability, and support for research tasks.

Topics of Interest

We encourage the submission of papers covering, but not limited to, one or more of the following topics:

• Representation
  • Data models for the description of scholarly data and their relationships, including rich semantic representations of hypotheses, methods, claims, and research results.
  • Description and use of provenance information of scientific data.
  • Integration and interoperability models of different data sources, including cross-domain and interdisciplinary knowledge alignment
  • NLP and AI approaches that demonstrate related methods and technologies.
  • Relevant knowledge graphs and ontologies.
  • Hybrid or LLM-based approaches for representation and knowledge graph engineering.
  • Infrastructures and metadata standards aligned with the Barcelona Declaration to ensure open and sustainable research information.
  • Applications of representation approaches in scholarly communication, including semantic publishing and structured scientific communication.
• Discoverability
  • Methods for extracting metadata, entities and relationships from scientific data.
  • Methods for the (semi-)automatic annotation and enhancement of scientific data.
  • Methods and interfaces for the exploration, retrieval, and visualisation of scholarly data.
  • NLP and AI approaches that demonstrate related methods and technologies.
  • Hybrid human-AI workflows for discovery, including curation, validation, and knowledge refinement.
  • Methods supporting interdisciplinary discovery and cross-domain knowledge exploration.
  • Applications and use cases demonstrating how ontologies, knowledge graphs, and LLMs support research tasks, such as literature exploration, hypothesis generation, and knowledge synthesis.
• Assessment
  • Novel methods, indicators, and metrics for quality and impact assessment of scientific publications, datasets, software, and other research output.
  • Uses of scientific knowledge graphs and citation networks for the facilitation of research assessment.
  • Studies regarding the characteristics or the evolution of scientific impact or merit.
  • NLP and AI approaches that demonstrate related methods and technologies.
  • Approaches to research assessment aligned with responsible research evaluation initiatives (e.g., DORA, CoAra). .
  • Metrics and frameworks for evaluating the quality, completeness, and reliability of scientific knowledge representations, including knowledge graphs and ontologies.
  • Evaluation of discovery systems and AI-driven tools, including their effectiveness, transparency, interpretability, and support for research tasks.
  • Benchmarking and evaluation methodologies for scholarly data infrastructures and AI-based research support systems (using ontologies, LLMs, KGs).

Submission Guidelines

Submissions are welcome in the following categories:

• Full research papers (up to 12 pages + unlimited pages of appendices and references)
• Short research papers (up to 6 pages + unlimited pages of appendices and references)
• Vision/Position papers (up to 6 pages + unlimited pages of appendices and references)

The workshop calls for full research papers, describing original work on the listed topics, and short papers, on early research results, new results on previously published works, demos, and projects. In accordance with Open Science principles, research papers may also be in the form of data or software papers (short or long papers). Data papers present the motivation and methodology behind the creation of data sets that are of value to the community, e.g., annotated corpora, benchmark collections, and training sets. Software papers present software functionality, its value for the community, and its application. To enable reproducibility and peer-review, authors are requested to share the DOIs of datasets and software products described in the articles.

The workshop also calls for vision/position papers providing insights towards new or emerging areas, innovative or risky approaches, or emerging applications that will require extensions to the state of the art. Vision papers do not necessarily have to present results but should carefully elaborate on the motivation and ongoing challenges of the described area.

Sci-K will adopt a single-anonymous review process and each paper will be reviewed by at least three Program Committee members.

Submissions must be in PDF format and must adhere to the CEURART single-column template. Submissions that do not follow these guidelines, or do not view or print properly, may be rejected without review. You can download an offline version with the style files from http://ceur-ws.org/Vol-XXX/CEURART.zip. It also contains DOCX template files. Overleaf users may want to use the CEURART template available in Overleaf. Please adhere also to the CEUR-WS Policy on AI-Assisting Tools.

The proceedings of the workshops will be published on CEUR (indexed in Scopus, DBLP and so on.)

Submit your contributions following the link: https://sci-k.github.io/2026/#submission

Important Dates

Paper submission: July 24th, 2026 (23:59, AoE timezone)
Notification of acceptance: August 21st, 2026
Camera-ready due: September 13th, 2026
Workshop day: October 2026 (to be confirmed)

Organisation

Allard Oelen, TIB Hannover, DE
Anna Jacyszyn, FIZ Karlsruhe, DE
Andrea Mannocci, CNR-ISTI, IT
Francesco Osborne, The Open University, UK
Georg Rehm, DFKI, DE
Angelo Salatino, The Open University, UK
Sonja Schimmler, TU Berlin, Fraunhofer FOKUS, DE
Lise Stork, University of Amsterdam, NL

More information about Sci-K 2026 is available at https://sci-k.github.io

Close Call for papers