Page 100 - AI for Good-Innovate for Impact
P. 100

AI for Good-Innovate for Impact



                      Use case – 21: AI for accessible and sustainable virtual worlds (for

                      education) (user needs and preferences)











                      Country: Spain

                      Organization : Universitat Autonoma de Barcelona.

                      Contact person: Estella Oncins (estella.oncins@ uab .cat)

                      Anna Matamala (anna.matamala@ uab .cat)

                      Pilar Orero (pilar.orero@ uab .cat)

                      Sarah McDonagh (sarahanne.mcdonagh@ uab .cat)


                      21�1� Use case summary table

                       Domain            Education

                       The Problem to be  •  Lack of cross-platform interoperability when transitioning between
                       addressed            different metaverses.
                                         •  Absence of consideration for user needs in AI-based education.

                       Key aspects of the  •  Integration of AI within the metaverse environment.
                       solution          •  Mapping various user interactions and providing guidelines to
                                            address anxiety and mental health within educational systems.
                                         •  Enhancing interaction between communication agents in educational
                                            systems through AI technologies.

                       Technology        Educational systems in virtual worlds; User centric approach (user needs
                       keywords          and preferences); Metaverse; AI-based Blockchain solutions in 2D, 3D,
                                         immersive environments

                       Data availability   •  Assessing the availability of relevant data for mapping workflows
                                            within educational contexts.
                                         •  Evaluating the accessibility of data regarding the use of AI, along with
                                            understanding associated barriers and benefits.
                                         •  Examining the availability of platform-specific scenarios that prioritize
                                            user interaction.
                       Metadata (type of   Textual data, Numerical data, Categorical data, Image data and Video
                       data)             data
                       Model Training    •  Using reinforcement learning algorithms for optimizing AI models in
                       and fine-tuning      the metaverse.
                                         •  Employing neural networks to fine-tune models for better communi-
                                            cation in educational systems.

                       Case Studies      None





                  84
   95   96   97   98   99   100   101   102   103   104   105