AI system for early childhood language development
Problem: Early language development is a worldwide problem. Children from over the world live in different socioeconomic backgrounds, which impact their language development. Children in disadvantage home environment are more likely to develop language problems because of the lack of language input. In consequence, these children have a risk of developing language impairments.
Solution: We propose an AI solution to improve learning and development of language in children under 5 years. This solution is designed to boost vocabulary growth in children through the interaction with an AI system.
Impact:
-Economic growth: children develop high cognitive abilities
-Social interactions: children will develop high communication skills
-Increase education outcomes
Early language development is crucial for children’s success in school and beyond. Children who develop strong language skills are more likely to arrive at school ready to learn. As a result, it is necessary to provide children with interactions that can support their growth and development. However, children from disadvantaged backgrounds(eg. abuse, violence, poverty, parents have lower levels of education) are more likely to have difficulties learning language and are more likely to have lower levels of achievement in school.
Children from different socioeconomic background will benefit from this solution. Particularly, we will focus on children in disadvantage home environment, because they are exposed to low language input. We will provide them with an AI system that interact and communicate with them everywhere and anywhere.
In our project, we're proposing an AI intelligent system to support the
early language development in children. This system interacts and communicate with children, so they will:
-Have direct access to information and learning.
-Learn conceptual knowledge.
-Increase their vocabulary.
-Learn name of objects and their features.
-Learn more information about objects.
- Enable parents and caregivers to support their children’s overall development
- Prepare children for primary school through exploration and early literacy skills
- Pilot
- New application of an existing technology
We offer an AI system to boost early language development in children under 5 years from all social and economic backgrounds. This system will help children learn and develop their vocabulary through interactions (questions-responses).
Our solution is based on AI technology to build an ontological conceptual system to help early language development. The implementation of this system follows several steps:
Step1: Extract features from child speech published in CHILD corpora
Step2: Create ontological conceptual system that represent learned nouns and link between them at every age
Step3: implement an AI application based on this conceptual system to learn names of objects and their features to children
- Artificial Intelligence
- Machine Learning
- Big Data
Activities:
Research activities: Conduct research on early language development in children. future activities: develop an AI system that can interact and communicate with children.
Outputs: research on early language development result on a number of publications:
M. Maouene, J.C Maouene & K.Canada Letting clusters and paths emerge from early semantic hypernetwork structure of features and their nouns WILD - workshop on infant language development, June 20th, 2013 to June 22nd, 2013, Donostia -San Sebastián, Spain
Ontologie taxonomique basée sur l’analyse des concepts formels. Application à la langue Tamazight. M.Maouene, N. Arkoubi, A. Moussa
10 e Colloque Africain sur la Recherche en Informatique et Mathématiques Appliquées-CARI2010,Côte d'Ivoire, Yamoussoukro : 18 – 21 octobre 2010
Shared Features in Networks of Early-learned Nouns. T. Hills, M. Maouene J. Maouene A.Sheya, , & L. Smith . Cognition Volume 112, Issue 3, September 2009, pp. 381-396, 2009
Early Semantic Networks: Preferential Attachment or Preferential Acquisition T. Hills, M. Maouene, J. Maouene, A. Sheya, & L. Smith. Psychoscience, Volume 20, Number 6, June2009 , pp. 729-739, 2009
The body region correlates of concrete and abstract verbs in early child language J. Maouene, N. Sethuraman, A.Laakso, M. Maouene 2011
Cognition, Brain, Behavior. An interdisciplinary journal copyright ©. Volume xv, no. 4 (December), pp.295-316
Short term outcomes: Extract features from CHILD data. Develop an AI system for English words and their features.
Long term outcomes: Extend our AI system for Tamazight language to help children in rural areas in Morocco.
- Children and Adolescents
- Infants
- Minorities/Previously Excluded Populations
- Morocco
- Morocco
Actually, this solution serves researches working on early language development. In one year, we would like to augment the number of laboratories worldwide and in Morocco particularly. In 5 years, our goal is to develop an AI system that will serve children in rural regions in Morocco ( Tamazight language)
In 1st year:
our goal is to work on existing CHILD data (English language).
In the next five years:
-Develop our AI system
-Run a psychological experiment on children in rural regions in Morocco to collect data and publish them in CHILD database.
-work on these data to improve language development in rural regions in Morocco
-Develop an AI application that can serve laboratories, child center, schools and cognitive psychologists.
- Lack of competencies in informatics and cognitive psychology.
-Networking: problem in establishing connections and partnership with specialists and social organizations.
-Some parents are unaware of the value of communicating with their children to improve their language skills, especially in disadvantaged areas of Morocco.
- Non homogeneity in the language used to speak with their children :
In mountain (also some rural ) areas, Tamazigh language is used
At home: Moroccan dialect ( DARIJA)
At School: Arabic/French
- Grow our team by selecting skilled people in informatics and software development
- Organize campaigns in mountain and rural areas of Morocco
- Contact child's centers, organize seminars and workshop
- Not registered as any organization
two persons
We're kaoutar skiker and Mounir Maouene, two members both interested in developing an AI system to improve the language development in children. Our ambition to is to to make our idea really and come up with an AI system to improve children language developement all over the world.
Mounir Maouene is a full professor in informatics and computer science with a an expertise in Artificial intelligence. He worked on child language development since 2005 . He has many publications ( See section More about yout solution).
Kaoutar Skiker has a PhD in computational and cognitive neuroscience. During her thesis entitled " The neurosemantic networks of object nouns: analysis and Intrpretation" supervied by Dr Mounir Maouene, she developed a system to assess the semantic memory system in people with Alzheimer diseases. Her current research work lies principally in the development, growth and breakdown of the semantic memory system in children, adults and dementia, respectively.
Our business model can be summarized as follow:
We will provide:
- AI system to promote early language development of children in difficult socioeconomic backgrounds. It will has a consequence on their late language and cognitive abilities.
These will benefit:
-Children in poor home environment.
-Labs working on early language development.
-Parents
-Social organizations
-Healthcare services
-Educational institutions
-Our product: develop our AI application and attract investors to move our social business forward.
In this way, our project will impacts:
-Research development
-Grow our social business
-Increase the effectiveness of child's centers
-Prepare children for a great future
-Long term consequence on overall society
Research investment:
- conduct research on children with neuro developmental disease such as autism.
Our revenue model can be schematized as follow:
To start our social business by applying to grants. It will help us to move forward and encourage us to grow up.
The next step, we sell our AI application to child centers, clinicians, IT organizations to finance our social program.
We are applying to solve to present out AI project to the solve community, which could provide us with invaluable mentorship and support. It will enable us to advance our research activities, grow our social business, increase our team, develop partnership to help us grow inside Morocco and beyond.
- Business model
- Technology
- Funding and revenue model
- Monitoring and evaluation
We would like to collaborate with laboratories working on early language development, social organization that are directly connected to children, cognitive psychologists, therapeutists, and child centers.
We plan to use AI technique to extract features from children speech available on CHILDE corpora. This extraction is important to create out ontological conceptual system and implement our AI application to promote language development in children under 5 years.
Indeed, this AI prize will provide us with a finantial support to:
- grow our team by hiring enginneers,sales, software developer,
- cover our travel cost to attend meetings and conferences,
-improve our skills in AI advanced techniques through training programs.