Dynamic, Real time, On-demand Personalisation for Scaling (DROPS)

Lead Research Organisation: University of Warwick
Department Name: WMG

Abstract

Privacy is a human right and value that often under threat in the context of technology. While much research has been done on this topic, in the modern era of technology privacy preserving technologies must keep in pace with state of the art technological developments, as in the case of dynamically personalised technologies. Personalised technologies operate on the basis of collecting personal data, dynamically analysing it to infer knowledge about the user and providing in-the-moment responses that add value for the user. The privacy implications of this technology, however, have been not explored extensively. DROPS will address this through an examination of the privacy, trust and identity issues that arise from the development of personalized e-books for children's reading. Our focus on children's reading is motivated by evidence that shows the value of personalised e-books for learning and reading enjoyment, and yet the lack of research that engages with the range of privacy issues that these technologies introduce.

DROPS has two broad synergetic goals: i) To develop a ThingsSpace that manages the computational process of personalizing technology, such as e-books, in such a way that a user's personal data is protected. ThingsSpace will use an existing personal data store, the HAT, to store personal data subsequently used by the e-book's algorithm to adapt to the child's pedagogical needs ii) To use ThingsSpace as a springboard to document and evaluate the privacy, trust and identity issues resulting from its design and use. This evaluation will lead to the development of different economic and business models creating a feedback loop with the technical design.

Across all of the project stages, we will co-create the technology with future end users of ThingsSpace (publishers, SMEs, designers of personalized digital products) and end users of personalised e-books (teachers, parents, children). This multi-stakeholder approach will allow us to ensure that the aspired values of e-books (for learning and privacy) are aligned with the economic models used to monetise this technology and the technical platform that delivers it. The code for ThingsSpace will be open source in order to allow other researchers and commercial parties to transfer our project findings to the diverse domains that digital personalisation is currently used, e.g. Finance, HealthCare, Social Media.

Planned Impact

DROPS focuses on the case of e-books with teachers and parents serving as the gatekeepers of these digital products who are often not cognisant of the privacy risks these complex technologies raise for young people who cannot yet give consent. Through co-creation workshops with these key stakeholders we will develop their understanding of how personalised technologies operate and the privacy issues that arise. This participation is expected to lead to preparedness that shapes their personal and professional practice.

We will develop novel technology by combining the latest state-of-the-art technology in personal data storage (the HAT) with privacy requirements in an emerging technology area, dynamically personalised digital products. The project will create new design methodologies and techniques that ensure there is alignment between good pedagogy, user experience and privacy in e-books. This knowledge will be disseminated and transferred through directly working with publishers of e-books showcasing the viability of our technology to their existing offering. Through this collaborative engagement with the commercial sector, we will also develop novel approaches to the economic models that take into account both privacy and financial viability.

We are committed to developing the skillset of future technology designers, in the domains of privacy and digital learning. We will develop projects and teaching materials across the four institutions involved in DROPS to develop know how and expertise in our undergraduate and graduate cohorts.

Public engagement will capitalise on the interdisciplinary partnerships and involve knowledge exchange activities such as workshops with children's media industry and publishers, high-profile media coverage, student projects and placements and relevant national and international conferences and scholarly publications.
 
Title Look and Feel Prototype of the Little DROPS children story creator/reader 
Description This is the creaive prototype of the UI of the Little DROPS children story creator/reader 
Type Of Art Artefact (including digital) 
Year Produced 2020 
Impact The development of the Little DROPS software application is dependent on this artistic product. 
URL https://invis.io/XHW88N134YP
 
Description HATLab/EPSRC DROPS project response to UN call on children's rights in the digital environment
Geographic Reach Multiple continents/international 
Policy Influence Type Participation in a national consultation
URL https://uwe-repository.worktribe.com/output/849564
 
Title BusyTime SHE Function 
Description Smart Hat Engine (SHE) functions are tools that are made open to those working on Hub of All Things (HAT) developers and can be accessed in the HAT sandbox. BusyTime is a function that analyses a GoogleDiary, looking at how often a person has appointments vs free time. It then returns a percentage 'busy time'. 
Type Of Technology Webtool/Application 
Year Produced 2020 
Impact We do not yet track the use of the SHE functions. 
URL http://stockvirtual.info/BusyTimeDemo.php
 
Title Little DROPS children story creator/reader 
Description Little Drops app concept has been born out of a series of design workshops with the participation of these industry partners as well as a few parents of young children. The app aims to facilitate the creative process of oral storytelling at home by providing a 'story template' and visual 'prompts' as supportive structure. Parents, children and their extended family can collect photographs from their everyday lives. Collated in a gallery accessible to the parent and child, these photographs can provide later inspiration by serving as 'prompts' to crafting a new story. Prompts are in turn utilized to populate 'story templates', which are based on spines of different story genres. Each spine is made out of blocks that support the story sequence allowing the parent and child to keep a cohesive narrative in their story. They also have the facility to record each story as it is told, allowing the parent to build a rich repository of the moments they shared with their children supporting reminiscence. 
Type Of Technology Webtool/Application 
Year Produced 2020 
Open Source License? Yes  
Impact The impact from the development of this software is not yet known at this time. 
URL https://github.com/mudspot/little-drops
 
Title Sentiment History SHE function 
Description Smart Hat Engine (SHE) functions are tools that are made open to those working on Hub of All Things (HAT) developers and can be accessed in the HAT sandbox.Sentiment History is a rule based function that counts the number of positive and negative words that appear in a given text. If the number of positive word appearances is greater than the number of negative word appearances, the system returns a positive sentiment, and vice versa. 
Type Of Technology Webtool/Application 
Year Produced 2019 
Impact We do not yet track the use of the SHE functions. 
URL https://www.hat-lab.org/sandbox
 
Title WordCloud SHE Function 
Description Smart Hat Engine (SHE) functions are tools that are made open to those working on Hub of All Things (HAT) developers and can be accessed in the HAT sandbox. WordCloud is a function that analyses text data, counting how often a particular word appears, giving it a ranking. The output can be used in visualisations such as Wordle, usually with the font size linked to the words predominance. The tool can also indicate preference etc. based on popularity of words. 
Type Of Technology Webtool/Application 
Year Produced 2018 
Impact We do not yet track the use of the SHE functions. 
URL https://www.hat-lab.org/sandbox
 
Description Improving productivity in Hollywood using data science 
Form Of Engagement Activity A broadcast e.g. TV/radio/film/podcast (other than news/press)
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact YouTube presentation of our paper "Improving productivity in Hollywood with data science: Using emotional arcs of movies to drive product and service innovation in entertainment industries". This video is used to share the paper with students and general public. It also helps engage academics, and I get invited to do the talk at seminars.
Year(s) Of Engagement Activity 2020
URL https://youtu.be/rfa66Lv_Lzs
 
Description Written piece about project findings on shared reading 
Form Of Engagement Activity Engagement focused website, blog or social media channel
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact The aim of the blog was to communicate the findings from a study involving 7 families and their shared reading practices. The blog was sent to the participating families and to new potential participants for further recruitment which has since materialised. It led to further discussions with early career researchers at UCL's knowledge lab, one of whom subsequently drew on the methodology reported in the blog for their own research.
Year(s) Of Engagement Activity 2019
URL https://medium.com/hub-of-all-things/personalised-digital-tech-for-shared-reading-9ebdcc0c6565