Wikipedia:GLAM/ARC Centre of Excellence for Automated Decision-Making and Society

The ARC Centre of Excellence for Automated Decision-Making and Society (ADM+S) is a cross-disciplinary, national research centre, which aims to create the knowledge and strategies necessary for responsible, ethical, and inclusive automated decision-making.

The ADM+S Centre recognises that to increase general understanding of automated decision-making technologies, applications and impacts working with Wikipedia, Wikidata and other Wikimedia platforms is one of the most effective approaches. To support this process ADM+S has established a Wikimedian-in-Residence position to support engagement between Wikimedia platforms and the researchers, students, partners, projects and issues associated with the Centre. This is the first time an Australian university research centre has hosted a WiR, with RMIT University and ADM+S Research Fellow User:AmandaSLawrence starting in the part-time (around 0.2 or one day per week) role in July 2021.

The ADM+S WiR role has involved a range of activities over the 2021-2022 which have focussed on 6 main areas:

• Awareness: Increase awareness and engagement with Wikimedia platforms by ADM+S researchers
• Training: Provide training for ADM+S researchers to become editors
• Content: Improve coverage of ADM+S topics on Wikimedia platforms
• Profile: Explore ways of representing and linking ADM+S people, partners and publications through Wikidata
• Analysis: Support critical thinking about Wikimedia as a knowledge commons and digital platform
• Exchange: Be a knowledge broker between Wikimedia and the centre on activities, projects and advocacy

Presentations

Worlds of Wikipedia Conference, Diversity, 'Diaspora and Digitality: discover the Worlds of Wikimedia and Beyond', 17-18 November 2022, Sydney University. Presentation on Wikimedia and the evidence ecosystem.

Wikipedia and Education in the Time of the ‘Crisis of Information’, Symposium, 17 September 2022, University of Canberra. Presentation on Wikipedia and reliable sources.

Automating Wikimedia: Open Knowledge, Linked Data and Search Seminar, Online, 4 May 2022, ADM+S.
Speakers: Assoc Prof Heather Ford: Head of Discipline for Digital and Social Media, UTS, Mr Liam Wyatt: Senior Program Manager, Wikimedia Foundation, Prof Mark Sanderson, Professor of Information Retrieval, RMIT University and Chief Investigator, ADM+S; Prof Julian Thomas, Director of the ARC Centre of Excellence for Automated Decision-Making and Society, and a Distinguished Professor in the School of Media and Communication at RMIT University. MC: Dr Amanda Lawrence, Research fellow & Wikimedian in Residence, ADM+S.
In this ADM+S event, panelists discuss how the power of search engines has become further entrenched in the wake of the current move to restructure the Web according to the logics of ‘linked data’ and the ‘semantic Web’. With the goal of sharing information according to structured formats that computers (rather than humans) can easily process and analyse, linked data engineers are developing algorithms that abstract information from fact sharing websites like Wikipedia into short, uniform statements that can be more efficiently shared, compared and analysed. Yet it is often unclear when viewing structured information on search results where the information is coming from – some is from Wikimedia but not all – and whether it is credible. At the same time the Wikimedia Foundation has launched Wikimedia Enterprise a new cross-departmental service of the Wikimedia Foundation available via enterprise.wikimedia.com which aims to provide commercial services for ‘organizations that want to repurpose Wikimedia content in other contexts, providing data services at a large scale’. What are the implications of these recent trends for verifiable search results, fact checking and for the free knowledge movement? Watch the Seminar video recording on Youtube

WREN: Reviewing A Year of Wikimedians in Residence Activity, Wikimania August 2022, Video -

Training

ADM+S Wiki Wednesday Drop-in sessions 2022: A chance for ADM+S researchers to learn more about how to work with Wikipedia, Wikidata and other Wikimedia platforms.

ADM+S Editathon October 2021 10.00 am - 1.00 pm (AEST) Wednesday 27 October 2021
En edit-a-thon to bring together ADMS members and other interested people to work on Wikipedia (and Wikidata) content related to automated decision-making systems and their social, legal and ethical implications.

Content Editing

Automated decision-making systems are not purely technological: they comprise a number of elements, including the data that enables predictive decision-making machines; the institutions that provide regulatory settings; and the people who design, use, imagine and adapt automated systems, and whose lives are affected by them. The centre is looking at these four elements across four domains: news and media, health, social services and transport and mobility.

There is already of lot of content on Wikipedia about technology, automation, AI etc but many pages are flagged for improvement and many do not adequately cover the social impact (whether positive or negative) and ethical issues, which is something ADM+S would like to address. Outline of Artificial Intelligence provides an overview of the content on Wikipedia relating to AI and Outline of Automation provides an overview of pages relating to automation. There are also topics, organisations and people missing - either within existing pages or that require new pages to be created. Citations to quality research sources are also a key issue for Wikipedia content.

The WiR role has involved creating new Wikipedia and Wkidata pages and editing many others:

• Algorithm
• Automated decision-making - new page created
• ARC Centre of Excellence for Automated decision-making + Society - new page created
• Crowdsourcing
• Data sovereignty
• Digital inclusion - new page created
• Digital divide
• Domestic violence - use of technology
• Evaluation measure (Information retrieval)
• Explainable artificial intelligence
• Information asymmetry
• Information retrieval
• Online advertising
• Recommender systems
• Sarah Pink - new page created

Editing and improving a selection of topic pages on Wikipedia might be a way build a collection of keywords in ADM, reflecting Raymond Williams well known collection of essays Keywords and other projects inspired by this such as the NYU Press Keywords series.

Wikidata

Work on Wikidata has included adding all the ADM+S researchers and partner organisations and related information as well as various technologies and companies of interest. Australian Research Council grant ids were added as a property to Wikidata which is useful given Wikidata is an important source of verification and linked data on funding for researchers, institutions and projects. Wikidata Query Service can now be used for visualising the networks of relationships across organisations that reflect the work of the centre and interest in ADM in Australian and across the world.