July 30, 2020, by Hazel Sayers
Smart Products Beacon students – update (1)
Rebecca Gibson has been working within the Hybrid Gift project to explore how hybridity in products can develop new gifting experiences. Her research has involved collaborating with (a large company) and working with other PhD students on the development of new hybrid gifts, which has involved meeting with stakeholders, planning and running a study to trial hybrid gifts and presenting back her findings. In addition to being a valuable experience from which she has gained insight from other PhD students along with the opportunity to work with a large company, Rebecca feels that being part of the Smart Products Beacon has provided the additional benefit of working alongside experienced researchers towards a shared goal of exploring digital technology within products, to generate new, exciting experiences.
Gisela Reyes Cruz has been working on I-CUBE, in the analysis of audio-visual recordings. Some experiments were conducted, in which participants had to instruct a “robot” where to put different items of clothing, in a predefined set of baskets. The robot in these first experiments was an actor or actress. “It’s been very interesting how people reacted to and instructed a person playing the role of a robot”.
Gisela’s involvement has been in the analysis of speech – everything participants said during the experiment. She has been revising and correcting existing transcriptions of the recordings, which, she says, has given her a good glimpse of the accuracy of automatic transcription services available to researchers in the University.
Gisela has also been coding the speech in the transcriptions, looking at how the instructions were structured, the words used to describe items, locations and other particularities in the interactions between participants and robot. She told us:
“Besides the insights coming from the analysis itself, being part of this project has led me to experience how research teams are organised and how to better manage different types of data of a large pool of participants”.
Ana Rita Pena has been continuing to work on her PhD exploring Fair Privacy-Preserving Insights. In this stage of her project Ana has been generating synthetic data sets and creating simple credit risk assessment toy models, to begin her exploration of the impact said models in different subpopulations. Initially she is exploring gender specific models and gender unaware models. She has created linear and logistic regression models for the data sets and is currently working on a privacy enhancing logistic regression ( differentially private logistic regression). In the near future Ana will start creating models based on decision trees and their privacy enhancing versions.