4 June 2024 in Journal of Documentation
Purpose: Community-generated digital content (CGDC) is one of the UK’s prime cultural assets. However, CGDC is currently “critically endangered” (Digital Preservation Coalition, 2021) due to technological and organisational barriers and has proven resistant to traditional methods of linking and integration. The challenge of integrating CGDC into larger archives has effectively silenced diverse community voices within our national collection. Our Heritage, Our Stories (OHOS), funded by the UK’s AHRC programme Towards a National Collection, responds to these urgent challenges by bringing together cutting-edge approaches from cultural heritage, humanities and computer science.Design/methodology/approach: Existing solutions to CGDC integration, involving bespoke interventionist activities, are expensive, time-consuming and unsustainable at scale, while unsophisticated computational integration erases the meaning and purpose of both CGDC and its creators. Using innovative multidisciplinary methods, AI tools and a co-design process, previously unfindable and unlinkable CGDC will be made discoverable in our virtual national collection.Findings: There currently exists a range of disconnected, fragile and under-represented community-generated heritage which is at increasing risk of loss. Therefore, OHOS will work to ensure the survival and preservation of these nationally important resources, for the future and for our shared national collection.Originality/value: As we dissolve barriers to create meaningful new links across CGDC collections and develop new methods of engagement, OHOS will also make this content accessible to new and diverse audiences. This will facilitate a wealth of fresh research while also embedding new strategies for future management of CGDC into heritage practice and training and fostering newly enriching, robust connections between communities and archival institutions.
22 April 2022 in Natural Language Engineering
Recent years have seen a growing number of publications that analyse Natural Language Understanding (NLU) datasets for superficial cues, whether they undermine the complexity of the tasks underlying those datasets and how they impact those models that are optimised and evaluated on this data. This structured survey provides an overview of the evolving research area by categorising reported weaknesses in models and datasets and the methods proposed to reveal and alleviate those weaknesses for the English language. We summarise and discuss the findings and conclude with a set of recommendations for possible future research directions. We hope that it will be a useful resource for researchers who propose new datasets to assess the suitability and quality of their data to evaluate various phenomena of interest, as well as those who propose novel NLU approaches, to further understand the implications of their improvements with respect to their model’s acquired capabilities.
28 October 2024 in Engineering Applications of Artificial Intelligence