Language Jungle - A Multimodal Active Projection for Facilitating Language Acquisition in Early Childhood -
Our idea aims to help parents facilitate language acquisition in children by offering multimodal linguistic interactive environments.
Updates: How has your idea changed or evolved throughout the Prize? What updates have you made to this submission? (1500 characters)
Our idea grew significantly throughout this Prize mainly by user centered design and mentorship feedback.
As parts of user centered design, we have iteratively conducted prototyping and user tests throughout the Prize. We have found a lot of hints at each user test to refine the prototype (e.g. to draw attention of child during usage of our product, to increase the retention time, to increase the number of utterances of child).
Regarding the mentorship feedback, we have learnt mainly the following things from our mentor:
- US’s social economical situation relevant to early childhood development
- Curated design to make user experience more inspiring
- How to make a compelling storytelling
- How to set KPIs of user test and what kind of information to be shared with parents to receive feedback
We have reflected some of the feedback to the prototype so that we could observe how users react differently.
Name or Organization
Tokyo/Nagoya, Japan. Target areas will be US, Japan and regions with similar problem in future.
What is your stage of development?
New Innovator, with less than one year of experience in ECD
What is the stage of your proposal?
Research & Early Testing: I am exploring my idea, gathering the inspiration and information I need to test it with real users.
Describe how your solution could be a game-changer for your selected Opportunity Area (600 characters)
A gap exists in the home environment for childhood language acquisition. Some parents can afford to hire a professional babysitter who could make positive impact on language acquisition. Some have sufficient time and knowledge to engage in desirable linguistic interaction with their children. But others don’t. Closing this gap is a massive problem to solve. Existing solutions need a labor-intensive approach and consideration of parents’ socioeconomic situation. Our solution could help the have-nots by partially automating linguistic interactions with their children and raising their awareness.
Select an Innovation Target
Product: A new or enhanced physical product that creates value for end beneficiaries.
Tell us more about your innovation (1500 characters)
Our innovation is a product to address the problem of the gap seen in home environments for children's language acquisition in the U.S. and throughout the world (Suskind, 2015). The children of the "have-nots" parents would have more linguistic interaction opportunities using our product. Language Jungle’s design was based on the “guided play” concept (Hirsh-Pasek et al., 2009) utilizing voice and other modalities. Moreover, parents will have access to insights about their child’s language acquisition through usage of the product. The insights would be shared with parents after the usage so they could learn and adapt the way of raising their children.
What problem are you aiming to solve? (3 sentences)
Recent research (Suskind, 2015) suggests that less opportunities for linguistic interaction in early childhood lead to negative impact in a child’s intelligence.
We would like to address this problem seen in poverty class in American and Japanese parents who cannot afford to give sufficient linguistic interaction with the childs due to a lack of time, money, or knowledge.
Our solution is to create an environment with affluent multi-modality and interaction to facilitate language acquisition.
Explain your idea (5000 characters)
Language Jungle (LJ) is a product that creates interactive linguistic environments at home or facilities and facilitates language acquisition in early childhood. Image 1 explains overall product concept with required technologies and the scope of the current prototype.
LJ mainly consists of the following elements:
- a microphone to record a child’s utterances (see a1 in Image 2)
- an algorithm to analyze the recorded child’s utterances and determine the timing and the content of the feedback (b)
- display(s) (e.g. speaker, projector, tablet, toy) to give multimodal feedback (e.g. vision, hearing, olfaction) based on the analyzed result (a2, c)
The flow of LJ usage is as follows:
0. (If necessary) LJ or parent feeds some inputs to trigger the baby's utterance
1. The child gives utterance which is recorded with the microphone
2. Analyzed results of the child’s utterances are converted to effective outputs rendered on the displays in a multi-modal way
3. The rendered outputs are reflected to the child as feedback to facilitate further utterances
4. Audio feedback can also be given so that the child can learn from it
5. Insights gained through the linguistic interaction between the baby and LJ are shared so that the parent can raise their children.
Through interactions with LJ, the child is actively encouraged by multimodal feedback to vocally communicate with the world around them . Interactions directed by the child would create more opportunities for utterances and learning based on enriched feedback. Language and even self-efficacy are expected to be acquired eventually by using Language Jungle.
Who benefits? (1500 characters)
For parents, the advantage of introducing Language Jungle is as follows:
(1) to relieve their busyness
(2) to encourage more communication when raising their children
(3) to recognize and improve the developmental situations of their children
First of all, Language Jungle frees parents from their busyness, as children pay most attention to interacting with Language Jungle. Second, Language Jungle directs parents on how to talk with their children according to the child's developmental stage. Third, using data corresponding to a child's developmental stage, parents can change the style of engagement with their children.
For children, Language Jungle provides them more interactions and conversations. Children are afforded opportunities to develop their vocabularies and intelligence by interacting with Language Jungle.
For child caregivers, preschool teachers and babysitters, Language Jungle supports their work and facilitates improvement of their teaching methods.
For researchers, Language Jungle provides longitudinal data of children’s behaviour. They can get the effect and result of intervention by Language Jungle in the earliest stages of life.
What kind of impact will your idea have? (1500 characters)
At this point, from user test, only short-term outcome (e.g. retention time) was able to measured. However, as you can see from Video 1, the child experienced the user test newly understand the concept of "going to the left/right." We think this shows potential of long-lasting outcome.
In our concept, by receiving feedback from Language Jungle, parents can always grasp the linguistic state of their children and improve their involvement. That is, it is expected that daily interactions are improved as the days go by.
Also, as the usage data of Language Jungle stored as part of life-long data could be invaluable for research on early childhood development, our idea have potential to impact the science community positively by sharing the collected data.
How does or how could your idea impact low-income children? (1500 characters)
The proposed solution Language Jungle would have several variants from high-end model to low-end model. The low-end model could be implemented into a
smartphone app without using other devices like a speaker, which could be still affordable to lower-income families.
In addition, from business model perspective, government support or research program could be incoporated to distribute Language Jungle in an affordable way.
Innovation: What makes your concept innovative? (5000 characters)
Due to an emergence of enabling technologies such as voice recognition and machine learning, we speculate that technology would be able to complement human's task in some fields including early child development space.
Specifically speaking of language acquisition aid, current solutions are mainly taking labor-intensive costly approach (e.g. recruiting babysitter) or passive approach like just monitoring the number of parent's/child's conversations to grasp the situation quantitatively. Our solution would be able to intervene in language acquisition environment for child actively in a non-labor intensive way, which makes a differentiation with existing approaches.
Scale: Describe how your idea could reach a significant number of end-users. (1500 characters)
As we are a new stage innovator, we don’t have a concrete plan for scaling yet. But we estimated potential market size in Japan as 2 million family, considering the number of families in poverty class and double income families whose parents may not have sufficient time for linguistic interaction with the child.
Feasibility: Where are you with understanding the feasibility of your idea? Describe what you’ve done so far and your plans. (3000 characters)
As we are a new stage innovator, we haven't finished yet both feasibility check from technical perspective and viability check from operational perspective. However, as shown in Image 1, we set milestones check the technical feasibility step by step. Currently, in the prototyping phase, we are focusing on specifying technical requirement based on the result of the user tests.
Business Viability: How viable is your business model? (5000 characters)
As we are a new stage innovator, we don’t have a concrete business model yet. But, to address both families in poverty class and in non-poverty class who need of language acquisition aid, two different business model are under investigation:
When approaching to the former segment, we shall need to deliver it to the end users in an affordable way. We would like to incorporate government support to solve this social economical problem together. In addition, as collected usage data of Language Jungle can be invaluable for child development research, Language Jungle could be a part of reserach program and distributed as a tool to collect data for research.
When approaching to the latter segment, a simple business model, where Language Jungle would be sold as a commercial product and end users pay the fee in return, would be applied.
We will build more concrete business models and verify them for both US and Japan market as the next step.
HCD: How have you used human centered design to build or refine your concept? (5000 characters)
We understood word gap is a big issue especially among families in poverty class. However, as we thought there would be some non-lower income families in Japan may face with similar probrem, we started from clarification of the local parents’ problem at the beginning by investigation and observation. Image 3 shows an example of weekday time schedule of two incomes families who don’t have time for sufficient linguistic interaction. We found a white space exists where both parents can’t engage with child. Video 8 shows an example of the scene. When the child’s mother was at home but has to do some housework, child had to take care of himself, wanted to intrupts her mother, but made no linguistic interaction. From this observation, we came to think we could address to fill this white gap by offering our Language Jungle to complement language acquisition.
After that, prototyping and user tests have been conducted iteratively to refine the concept.
For prototyping, we developed a software, where we can switch language between Japanese and English to enable testing with different languages and customize the design quickly.
Regarding the user test, we chose one focus family, one family from our targeting segment, and one family who has 5 children as an extreme user.
Until the 3rd iterations, user tests have been conducted with the focus family only, so that basic features of Language Jungle can be tuned quickly. At the 4th iteration, user tests have been conducted with 3 families to verify individual differences and differences in perception when using different media (e.g. PC display, TV, projector)
And at last, at the 5th iteration, using the method of Wizard of Oz, we simulated user experience as if Language Jungle become smart enough for effective linguistic interaction, so that we can see if our idea has potential to be accepted at the end.
The insights we have obtained from user tests are summarized in the attached file "LJ_user_test.xlsx.pdf."
Tell us more about you (3000 characters)
MITSUGO project is not a company but a moonlight work with collection of individuals who have different expertise such as robotics, learning design, and learning science. As most of us have own child, we have strong passion to wok on issues in early childhood especially for language acquisition. When knowing the fact that “word gap” problem can’t be overlooked, we thought there may be some way to solve the problem at least partially using technology.
As it is quite natural for us to have user’s perspective against the problems in early childhood, we can come up with ideas and test them easily and quickly. We also have motivation to solve own problems or problems we can emphasize.
Our team owns technical competence at least for prototyping and domain knowledge in early childhood development. We have a robotics expert who had experiences to build interactive systems, a learning scientist who is studying about confirmation bias, conceptual change and knowledge system in childhood, and a learning designer.
Do you have the people and partners you need to do what you’ve described? (600 characters)
Yes, we will be part of Singularity University’s incubator program (i.e. SU incubator) in Silicon Valley from April 2018 and be based in US, where our current idea will be more concrete and fit to US market through conducting user research and market research.
As you consider your next steps, what kinds of help could you use? Is there a type of expertise that would be most helpful? (1800 characters)
To conduct market research and user research mentioned above, we still need partners who could support the following actions:
- Investigation of US’s social economic situation related to language acquisition in detail
- Access to targeting audience (e.g. child facilities or parents who has child from 0 to 3)
Would you like mentoring support?
If so, what type of mentoring support do you think you need? (1200 characters)
- UX experts who is experienced in designing UX for parents and children
- Developmental psychology
Are you willing to share your email contact information submitted on OpenIDEO with Gary Community Investments?
Yes, share my contact information
[Optional] Biography: Upload your biography. Please include links to relevant information (portfolio, LinkedIn profile, organization website, etc).
[Optional] Attachments: Please upload relevant attachments or graphics or show us how you prototyped.
Video 9 shows how our software works for user tests. The displayed object can be controlled both automatically and manually. At user test with prototype version 5, Wizard of Oz method has been applied to tune the parameters and to design how the system interacts with child.
Mentorship: How was your idea supported? (5000 characters)
Our mentor Mr. Zachary Snader brought about very positive impact to our concept and networking.
He shared US’s social economical situation relevant to early childhood development, so that we could clarify our target audiences in US.
Besides, thanks to his feedback on storytelling and KPIs for the user test , we realized some missing elements in our solution (e.g. parent’s vocabulary setting for the customization, target to be fulfilled).
In addition, he kindly introduced us to the other team who could potentially offer fields for user test in US.