openTEL is pleased to announce that next week’s openAIED Special Interest Group will be exploring the ethical implications of AI research projects and how machine learning can be used to automate the process of module mapping.
Towards an “Ethics by Design” methodology for AI research projects
Abstract: There is a rapid increase of AI projects and applications resulting in fresh debates over societal and ethical concerns. Addressing the potential risks emanating from AI is a great challenge due to the nature of research and the specific set of technical skills involved. AI and Data Science researchers are not trained to identify, anticipate and analyse such sociotechnical issues, or to establish solutions at the time a specific research project is being designed. Therefore, this talk suggests a methodology for an ethical research design that involves a broader set of skills from the start of the project and throughout its lifecycle. The methodology is inspired by design literature, urging for the inclusion of a broader understanding of ethical and societal risks emerging from AI.
Bio: Pinelopi Troullinou is a Research Associate at the Knowledge Media Institute (KMi) of The Open University, having an interdisciplinary academic background on social sciences and over nine years of research experience. Her research focuses on privacy, ethical and societal concerns of digital technologies and algorithmic governance. She is the lead scientist of the innovative workshop series ‘re:coding Black Mirror’, hosted already in two major international Computer Science conferences (International Semantic Web Conference in 2017 and The Web Conference in 2018) and to be hosted at Computers, Privacy and Data Protection (CPDP) 2019. Pinelopi has also served as guest co-editor for the special issue ‘Redesigning or Redefining Privacy?’ published by the Westminster Papers in Communication Culture (WPCC).
Automating module mapping using machine learning
Abstract: The Learning Design team at The Open University conducts a process known as ‘module mapping’ on new and redesigned modules. This process involves the categorisation of module activities into a taxonomy of seven categories and is laborious, typically taking between three and five days per module. This talk will present some proof-of-concept work using machine learning to automate this process for one specific category as well as discuss issues encountered and possible future directions.
Bio: Juliette Culver has worked as an Educational Technology Developer in the Institute of Educational Technology and Learning and Teaching Innovation (Academic) for the last 13 years. She has a doctorate in Mathematics from the University of Oxford and has worked at Royal Holloway, University of London and at the University of Surrey, as well as having experience working as a software engineer in the private sector.
We look forward to seeing you at 14:00 on 15th January in Library Seminar Room 1, Library.