DWW Research work

Project 1 | Project 2 | Project 3 | Project 4 | Deepening

Presentation

Subject: Provide easy access to learning in-demand skills

Disclaimer: in-demand skills are often defined as the skills that are the most wanted by recruiters on the market. In light of the 4th ODD, I made the choice to focus on in-demand skills that are the most important and basic in poor countries: reading and maths.

Technological module : Database

A. Exploratory Research

Project 1: MathsGee STEM Community

  1. Project carriers

    2. MathsGee is a non-profit organization start-up founded by Edzai Conilias Zvobwo in South Africa

    The plateform has a lot of other partner educative platforms such as: Blackboard, Brightspace...

  2. Beneficiaries
    • African students/pupils
    • African teachers
  3. Users
    • African students/pupils
    • African teachers
  4. Need
    • No easy access to education
    • No ressources available to buy or access online courses
    • Learn from everywhere and without the need of a lot of infrastructure
    • Access education without having to find a school
    • Africa does not have enough educative content creators
  5. Principle
    • Q&A platform : everyone answers and asks questions
    • Marketplace for teachers who can post free content or even sell courses more personally
    • It is also possible to take appointments with a tutor to get personalized help
  6. Technologies
    • Uses the solution sold by invision community for the Q&A system
    • Use of databases : users, questions and answers
    • Website in html, js, css
    • Use of jquery, google fonts, Bootstrap and Fontawesome
    • Google analytics, google tag manager, google ad service
    • Hotjar for behavior analytics and feedback data
    • Mathjax correct display of mathematics
    • Scripts written in PHP
    • Servers hosted by CloudFare and NGINX
    • Slick: Database query and access library for Scala
    • Disqus/tawk.to to interact with users
  7. Sources

Project 2: Tusome

  1. Project carriers
    • Kenyan Ministry of Education
    • RTI international, independent and non profit research institute whose goal is to "improve the human condition"
    • USAID Organisation
  2. Beneficiaries
    • Kenyan students/pupils
  3. Users
    • Kenyan students/pupils
  4. Need
    • 40% of primary school pupils can not understand their reading material
    • Goal of improving 6.7 million children's reading abilities
  5. Principle
    • Provides 26 million manuals and additional supplies to primary school pupils, helping to guarantee that each pupil gets his own textbook.
    • Marketplace for teachers who can post free courses or sell them
    • Stands out with "tablets with digitized learning material, comments and close following of the progress of the child for the teacher." Every child's performances are uploaded to a network system based on cloud.
  6. Technologies
    • Android tablets
    • Digitized textbooks and learning materials, downloaded on tablets to ensure everyone can study, even without a connexion
    • Recording system for progress and notation/feedback system for teachers for those who have the internet
    • Cloud service to save and store results with database used for storing results and user information
    • Papaya app for learning pronunciation
    • Growing database for NGOs and governements to assess the efficience of the system and collect data to improve the project
  7. Sources

Project 3: EI Mindspark

  1. Project carriers
    • Educational Initiatives : ed-tech company
  2. Beneficiaries
    • Indian students/teachers
  3. Users
    • Indian students/teachers
  4. Need
    • Vast education inequalities in India
  5. Principle
    • Large collect of data in India and broader for six years
    • Creation of the Mindspark platform : adaptative learning thanks to AI, tailored rythm and progression according to the level and the profile of the child. The platform also uses data to compare the child to others
    • Also helps teachers to better understand children and better help them, coaching is easier and improved
  6. Technologies
    • Big data: databases to store all results, trends, scores, errors or successes. Apprently, databases contain up to two billion data entries
    • Questions database / quizz : more than 45 000
    • AI to treat all this data and extract models and insights for other students
    • iOS and Android app, because the app is quite small (40 mb for the iPhone app), we could assume its a webapp
    • Website : favicon, api facebook for domain verification, google ad service, google tag manager, ms clarity library for behavioral analytics, Saas services from zoho for clients relations management, google fonts, jquery
  7. Sources

Project 4: Onebillion

  1. Project carriers
    • Onebillion: non profit organisation
    • A lot of NPOs for helping local implementation: Hello world, UNESCO and World Food Programme
    • Google
    • RTI international to supervise data and cognitive processes
  2. Beneficiaries
    • Children of the world who don't have access to education and especially to reading
  3. Users
    • Children of the world who don't have access to education and especially to reading
  4. Need
    • There is still a lot of illiterate in the world and especially in Africa
  5. Principle
    • They provide tablets with "onecourse" software for children to be able to learn. The tablet enables those who can not go to school to still get a proper education
    • Regional adaptation of the software: pronunciation and audios with local native speakers and adapted cultural elements
  6. Technologies
    • Android tablets donated by Google
    • Software: onecourse, installed for learning even without an internet connexion
    • Construction of a course database according to the level and the country
    • Collect of data to measure efficiency and the progress of the children (for those with internet access)
  7. Sources

B. Deepening

Selected project : EI Mindstark

  1. Project carriers
    • Educational Initiatives : entreprise de l'ed-tech
  2. Research question : How can technology help children who don't have access to education ?
  3. Why this project?

    4. As mentionned, children don't have access to a quality education but mostly to help and tutoring. Therefore, at first sight, I did not believe that technology was the answer. However, I believe that Mindspark's approach is very interesting regarding those issues: it can adapt to each student and contributes to solve this lack of tutoring and support but also the difference of levels in large classes.

User scenario

  1. Users

    2. Indian students and teachers

  2. Persona
    • Sridhar
    • 10 ans
    • He lives in the region of Haryana, poor rural region of India
    • Il loves to ride his bicycle and to play in the fields
    • He is very poor, he has to walk very long to go to school and often, he can not go because his parents need help. He has seven brothers and sisters and their house is too small for him to study properly. Later, he dreams of going to Dehli and working in computers. It is his dream but he does not even have a computer, he is already behind in maths and his reading does not progress.
    • His major issue: his education is not consistant enough and he does not get enough support
  3. Key features
    • Learn to read
    • Learn to write
    • Learn maths
    • Discover educational content
  4. UX Storyboard
    5. Photo du storyboard

Technical analysis

  1. General principle

    2. To sum up, the project is a learning platform as it is often seen: it hosts lessons, content et allows to follow online courses. However, the solution stands out by using AI. This AI, after 7 years of studying students, allows to create learning paths and progression paths for each student, according to his level and own understanding.

  2. Technical overview - Database version

    3. There is very few information on Mindstark's database as it is their way of stadning out. The only "official" information are the following: more than 5 billion data points and a questions/test database of more than 45 000 data points.

    Thus, because we don't really have in-depth information on the database, let's try and think about how they could work:

    • Questions database : +45 000 données
    • Students database
      • Name
      • Surname
      • Student number
      • Localization
      • Grade
    • Students' record database
      • Start date
      • End date
      • Time spent
      • Current level
      • Entry level
    • Events/logs database
      • Events
      • Logs

    We can then imagine examples for each action:

    • A data is created : a new student registers, a student completes a test
    • A data is read : a student completes a test, his results are compared to others to rank him so other grades are read
    • A data is updated : each year, every student's age is updated
    • A data is deleted : we can imagine that when a student finishes his education on the platform, his personal data are erased so that his performance is rendered anonymous for respect of privacy.

    Different types of queries:

    • Search : a student finishes a test, the AI searches for students who obtained the same score
    • Filter : to establish progression models for students, the AI can filter students with the same key parameters: progression, age, start date...
    • Sort : A points system is available for students to rank themselves amongst the others. A sort query can be used to sort students according to their total of points
  3. Added value thanks to database :

    4. Here, databases are central to the platform. In fact, AI implies a need of a large amount of data to be treated, well sorted and easily accessible. Therefore, each data is stored on a database and can be reused by the AI as a source of improvement and development of the student's tutoring. In this project, the whole added value lies on database and AI.

  4. Complementary source
    5. https://appadvice.com/app/ei-mindspark-teacher-app/1542215560