Leveraging AI technology to provide a unique and personalized storytelling experience for children.
Storytelling has been an integral part of human culture for thousands of years, and for children, it provides a fun and engaging way to explore the world and develop a strong sense of identity. Through storytelling, children can foster imagination and creativity, as well as build language and critical thinking skills. However, traditional reading and storytelling experiences may not always provide the level of engagement and personalization needed to capture children's interest and support their learning.
In this project, we aim to explore how we can leverage Artificial Intelligence (AI) and Machine Learning (ML) to provide a unique and personalized storytelling experience for children.
Course Project: Product Design | Experience Design| UI Design
Instructor: Prof. Tad Hirsh
Jane Effanga, Kristyn Weaver, Seung Hyun Rebecca Kim
Feb - Apr 2023
Figma, Miro, Illustrator
Use case research, Journey map analysis, Risk analysis, Data design, Mental model design, Onboarding experience design, Prediction and feedback experience design, Prototyping
How might we leverage AI/ML to provide a more engaging and personalized storytelling experience that supports the imagination and creativity of children?
My StoryWeaver is an innovative platform that blends the power of AI with the imagination of children. This concept aims to empowers young minds to co-create their own tales and shape the outcome of their stories, revolutionizing traditional storytelling.
My StoryWeaver aims to foster imagination and creativity in children while building their language and critical thinking skills. Our target audience is children aged 5-7 who are early readers.
How the story creation process works
The concept uses AI/ML algorithms to collaborate with children in generating story ideas, characters, and plot lines. The system prompts and suggests to guide children while giving them creative freedom. This approach develops their language and critical thinking skills through a fun and engaging experience.
Use case Illustration
Story element selection
The child chooses their preferred options from the list of story elements to initiate the story co-creation process
AI generates story outline, child chooses to start the story or start over with new story elements. Reading options are available for a personalized experience.
AI kickstarts story with intro and prompt, child responds, they can also use the "I don't know" button, which automatically fills in a response. And the AI continues the story based on response provided.
Throughout the story development process, the system alternates between prompts and multiple-choice options as a means of balancing engagement and cognitive load for the child. Additionally, an open-ended response option is included to allow for free expression of ideas.
The story-building process is designed as a continuous loop, but the child has the option to end the story at any time. At this point, they are presented with an ending and given the option to read the story, continue with the current story, or start a new one.
THE DESIGN PROCESS
✧ Use Case Research
✧ Data Collection
✧ Mental Model & Onboarding
✧ Predictions & Feedback
USE CASE RESEARCH
Validating our design concept
We conducted extensive use case research as the first step in designing the My StoryWeaver platform. Our objectives were to validate the design concept, understand workflows, optimize the AI's reward function, and identify potential outcomes and risks. This research involved a literature review and interviews with parents, children, and teachers. Through this, we identified key pain points and gained valuable insights into our target audience's needs and preferences.
Research activities and insights
Conducting validation research to assess the need and demand for the design concept.
Journey map analysis
Analyzing storytelling and story creation workflows to identify existing practices that can be augmented or automated by the concept.
AI reward function
Optimizing the AI reward function to incentivize engaging and relevant story prompts while maintaining narrative coherence
Expected outcome and risks
Anticipating outcomes and risks associated with the design concept.
Identifying the required data and sources
During the data collection process, we identified the specific data labels necessary for the model to deliver the intended user experience. This involved evaluating existing labeled and unlabeled datasets and identifying any new datasets that needed to be produced. For new or modified datasets, we developed a plan for how data would be collected and labeled, as well as any resources or incentives required for data collection.
MENTAL MODELS & ONBOARDING EXPERIENCE
Helping users understand how this works
We created a mental model for the My StoryWeaver project using the metaphor of a self-building train track. The model helps users understand how the process works and their role in directing the story. We also explored other metaphors, but ultimately chose the train track metaphor as the best fit for the project.
Illustration of My StoryWeaver using the metaphor of a self-building train track
Discovery and Pre-boarding
We considered various strategies for attracting young parents to My StoryWeaver during the discovery and pre-boarding phase. One of the strategies we implemented was advertising on social media and through Google search. When potential customers click on our ad link, they will be directed to our landing page, where our value proposition is clearly communicated. We emphasized the concept of co-creation in our messaging, highlighting the unique and personalized experience our platform provides for children and parents.
Account Creation Flow
Our goal was to design a streamlined onboarding experience that would collect all the essential information. One reason we asked for the parent's year of birth was to ensure their involvement in the account setup. Despite My Storyweaver being designed for children aged 5-7, we wanted to empower parents with some control and encourage their engagement in the process.
MAKING AND TUNING PREDICTIONS
Optimizing user experience through effective prediction communication and feedback Collection
Factors that lead to high and low predictions
High confidence scenarios
Low confidence scenarios
REFLECTION & FUTURE POSSIBILITIES
Throughout the project, our team faced challenges, including losing sight of the original concept of My StoryWeaver, which emphasized the magic of co-creation between children and AI. However, with the help of in-class critiques, we were able to refocus on the important aspects that we may have allowed to slip and identify the most effective components and automated features that could be incorporated into the platform.
Looking back, there were areas where the project could have been improved, including time management, user testing, and an in-depth analysis of the capabilities and limitations of existing generative AI technologies. However, despite these challenges, we are proud of the progress we have made in providing children with an engaging and inclusive story-building experience.
Moving forward, the next step is to conduct user testing with 50 families with children aged 5-7 and conduct post-usage interviews to address any issues with user interaction. With more time and resources, there is potential for My StoryWeaver to provide even more engaging and inclusive story-building experiences for children.