• Digital Humanities and Cultural Heritage
• Publications close
• Other research products close
• Instruct-ERIC close

This is a presentation given at the European Archaeology Association Conference 2023 in session 603 - Replicable Archaeology: looking for workflows and data management strategies fostering data reuse and methological transferability in archaeological science. 

Metadata is key to framing the context of the research data and can sometimes be part of the research data itself. The Metadata Standards Catalog (MSC), maintained by the RDA Metadata Standards Catalog Working Group, hosts more than one hundred descriptions of metadata standards, along with tools, mappings, and organisations that aim to help researchers choose the most suitable metadata standard to describe research data belonging to projects, repositories and catalogues. In the context of the Cultural Heritage domain, we examined the consistency and evaluated the MSC subject domains with a view to enriching the Catalog with the metadata standards widely used by the Cultural Heritage research community. In addition, we investigated whether the MSC could play a role in guiding researchers to include relevant citation information in their project metadata, disciplinary metadata and dataset metadata, thereby fostering cross-disciplinary information integration while proving, at the same time, how citation can act as a FAIR-enabling factor. Part of A FAIR-enabling citation model for Cultural Heritage Objects project activities. This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 101017536 and is supported by the EOSC Future (https://eoscfuture.eu/) through the RDA (https://www.rd-alliance.org/) Open Call mechanism (https://eoscfuture-grants.eu/provider/research-data-alliance), based on evaluations of external, independent experts.

The definition of the critical citation elements of cultural heritage objects (CHOs) informs one of the milestones of the project A FAIR-enabling citation model for Cultural Heritage Objects. It is based on data gathered through the examination of case studies focusing on CHO citation models in plural – native and non-native – contexts, the replicability of CHO citation as well as cross-disciplinary exploratory interviews, along with the identification of the critical citation elements within the CHO metadata schemas. This deliverable aims to illustrate the results achieved in terms of the distinctive features, criticalities, habits, and shortcomings in citation practices related to cultural heritage objects and data. Part of A FAIR-enabling citation model for Cultural Heritage Objects project activities. This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 101017536 and is supported by the EOSC Future (https://eoscfuture.eu/) through the RDA (https://www.rd-alliance.org/) Open Call mechanism (https://eoscfuture-grants.eu/provider/research-data-alliance), based on evaluations of external, independent experts.

Phytolith research contributes to our understanding of plant-related studies such as plant use in archaeological contexts and past landscapes in palaeoecology. This multi-disciplinarity combined with the specificities of phytoliths themselves (multiplicity, redundancy, naming issues) produces a wide variety of methodologies. Combined with a lack of data sharing and transparency in published studies, it means data are hard to find and understand, and therefore difficult to reuse. This situation is challenging for phytolith researchers to collaborate from the same and different disciplines for improving methodologies and conducting meta-analyses. Implementing The FAIR Data principles (Findable, Accessible, Interoperable and Reusable) would improve transparency and accessibility for greater research data sustainability and reuse. This paper sets out the method used to conduct a FAIR assessment of existing phytolith data. We sampled and assessed 100 articles of phytolith research (2016–2020) in terms of the FAIR principles. The end goal of this project is to use the findings from this dataset to propose FAIR guidance for more sustainable publishing of data and research in phytolith studies. This project has received funding through EOSC-life from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 824087. The FAIR Phytoliths Project is based at two institutions: Universitat Pompeu Fabra and Historic England. We would like to thank both these institutions for their support and also financial commitment to this project. Thank you also for in-kind support from Texas A&M University and the Spanish National Research Council for time on this project by Javier Ruiz-Pérez and Juan José García-Granero. Ruiz-Pérez was supported by the National Science Foundation under award number DEB-1931232 to J. W. Veldman. García-Granero acknowledges funding from the Spanish Ministry of Science, Innovation and Universities (Grant No. IJC2018-035161-I). Carla Lancelotti acknowledges funding from the European Research Council (ERC-Stg 759800). Celine Kerfant is a recipient of Juan de la Cierva-Formación contract FJC2021-046496-I funded by MCIN/AEI/10.13039/501100011033 and by the European Union “Next Generation EU”/PRTR.

The Gdańsk Tech Library has implemented a "Data Practices in an interdisciplinary perspective" project. The project results from an open call for interdisciplinary research adoption grants funded by the EOSC Future using the RDA Open Call mechanism. This project aims to create an overview and thematic analysis of open data curation and sharing within selected scientific discipline. The poster is based on a semi-structured interview with a researcher from the Gdańsk University of Technology to develop a case study for natural language processing. Collected data was analysed and documented graphically to illustrate the most critical procedures and requirements for data management for the particular discipline. The poster aims to describe the opportunities for data sharing, workflow practices in data management, implementation of metadata standards, and the social aspect of sharing data, such as researchers' attitudes and discipline-specific traditions. In addition, the poster includes a model example of a dataset for the natural language processing discipline deposited in the university data repository (MOST WIEDZY Open Research Data Catalogue).

Sentiment analysis is a vital tool for uncovering insights from financial articles, news, and social media, shaping our understanding of market movements. Despite the impressive capabilities of large language models (LLMs) in financial natural language processing (NLP), they still struggle with accurately interpreting numerical values and grasping financial context, limiting their effectiveness in predicting financial sentiment. In this paper, we introduce a simple yet effective instruction tuning approach to address these issues. By transforming a small portion of supervised financial sentiment analysis data into instruction data and fine-tuning a general-purpose LLM with this method, we achieve remarkable advancements in financial sentiment analysis. In the experiment, our approach outperforms state-of-the-art supervised sentiment analysis models, as well as widely used LLMs like ChatGPT and LLaMAs, particularly in scenarios where numerical understanding and contextual comprehension are vital. Comment: FinLLM Symposium at IJCAI 2023

The Social Sciences and Humanities communities needed a central point to gather and exchange information about their tools, services, and datasets. Although plenty of project websites, service registries, and data repositories existed, they were mostly fragmented. The lack of contextualisation connecting these assets and offering domain-relevant means to discuss and enrich them was evident. Also, there was a need to digitalise this research domain which, in most cases, was not yet digital native. This siloed environment was the challenge SSHOC set out to bridge in January 2019. The solution to solve this problems was found in the SSH Open Marketplace, delivered in 2022. It is a discovery platform supporting researchers in the multi-faceted domain of humanities and social sciences at every step of the research data lifecycle. The SSH Open Marketplace is available on the EOSC Portal and was developed by SSHOC The full list of EOSC in practice stories is available here

This document is a summary (brochure-style) document describing the Key Exploitable Results (KERs) of the Horizon 2020 EOSC-related projects still active as of the spring of 2022. The information was provided by the projects through a survey questionnaire that was developed in collaboration between the EOSC Association and the Research Data Alliance Association (RDA). The survey collected responses from 22 projects, which reported a total of 119 KERs, covering mainly technical and policy harmonisation efforts, virtual research environments, discovery/access platforms, training resources, knowledge centres and validation tools. The KERs identified in this survey of Horizon 2020 EOSC-related projects show strong correlations to the EOSC Advisory Groups (AG) topics, with maximum relevance for “Technical challenges in EOSC” and “Implementation of EOSC”; high relevance for “Metadata and data quality”; and satisfactory relevance for the AG topics “Research careers and curricula” and “Sustaining the EOSC”. The full report is available at: https://doi.org/10.5281/zenodo.7401539