Process Group 14

Return PRE2017 3 Groep14

Concept 1: Intelligent Quiz Master

Subject

Fun Learning for Kids Enhance knowledge levels of young children through an interactive quiz system. Teachers will be able to tell the system the desired final knowledge level and will be able to see the progress of each child. Furthermore the system has to be able to interact with the children and assess their knowledge levels to create questions on their personal level. In our project we will focus on the quiz and try to create this.

Users

• Children from class 3 - 4 in the Netherlands. The system could later on be changed to fit other age groups that have other knowledge levels, but for this project we focus on this group as the simple math they need to learn here are an easy starting point for the program.
• Parents or guardians of these children as they want to know the progress their child or children has made.
• Teachers that can tell the systems what level of knowledge the entire class needs to reach at the end. They also need to be able to see how far each student has gotten.

User Requirements

children:

• learning through a fun program
• competitive, want the highest reward (number of sheep)

parents:

• want their children to study properly
• want their children to be happy
• want their children to be motivated/interested
• want to have more free time as they do not have to tutor as much anymore
• want to check their children's progress

teachers:

• let the students learn effectively
• the system has to match the curriculum
• easier to check all students' progress

Objective

Develop a smart quiz program for on a computer/tablet/laptop that can assess knowledge levels of its users and ask questions on their personal boundary so they learn effectively.

Approach

Creating the smart quiz and interface in java.

Intelligent Quiz Master

Idea. Use a set of arithmic questions (addition, subtraction, fractions) since then it is easy for us to check if it makes sense. Also, since most children have difficulties with arithmic this is actually useful.

Given a set of questions, the quiz master will test the knowledge of a child, and help the child improve by asking the right questions at the right time. We will build an application that selects the next question to ask the child, based on the previous answers the child gave to previous questions. The quiz will find the knowledge level of the child and ask questions at the child's knowledge boundary so he can still learn from the question but will not be overwhelmed.

The quiz master has to:

• Find out the level of knowledge the child has, and ask questions that are on the 'edge' of a childs knowledge in order to improve their knowledge.
• Optionally invent new questions, similar to the already existing questions.

In order to do so, we must:

• Define distance (or question similarity) between questions, which questions are of similar difficulty. So cluster questions based on their difficulty. Note that this will vary per child.
• Simulate the (increasing/decreasing) knowledge of different children. (To be able to train our app.)
• Construct a (large enough) data set to use parts of it for training and validation.
• Find out what the next right question would be. Our app should do this, based on the question similarity for a certain child. Educational/psychological: what are the best questions to ask?

Milestones

• Finishing planning
• Summarizing SotA
• Quiz
  • Teacher can enter category and boundary (knowledge goal that has to be reached)
  • Quiz can generate questions inside category
  • Quiz can understand the person's input
  • Quiz uses input to generate personal level questions
  • Quiz gives results
• Interface
  • Results are displayed to child/parent/teacher (they each have their own interface)
  • Reward system (for the child, parents can see this as well)

Deliverables

Smart quiz program including interfaces for the child who will use the quiz to learn, the parents and the teacher.

State of the Art Literature Study

State of the Art Literature Study

+ UML for Quiz =

• This includes the categories and how we plan to make it smart.

Clustering

• Knewton did something similar...
• Ideas for defining distance between arithmic questions:
  • Number of operators.
  • Size of the numbers (number of digits, or a range).
  • Create graph of questions, with the weight of an edge between two nodes the number of similarities they have. An edge with weight 0 is not an edge. Then all components are their own cluster, and we can find minimal cuts in this network to construct more clusters.

Who will do what? Planning

• Abby focusses on the quiz programming
• Christine focusses on the quiz design
• Dennis focusses on the quiz programming
• Ellen focusses on the wiki
• Sophie focusses on the quiz programming

week 1:

• literature search, SotA summary
• make plan
• setup Git
• update wiki

week 2:

• quiz plan:
  • how to make it smart
  • which categories
  • UML of quiz
• update wiki

week 3:

• implement quiz structure
• quiz has to read input
• start with letting quiz learn from input
• update wiki

week 4:

• quiz has to learn from input
• start with interface for child
• feedback system in quiz
• update wiki

week 5:

• reward system in interface
• child interface has to be finished
• smart quiz has to generate questions inside the categories
• update wiki

week 6:

• teacher program finished
• parent program finished
• smart quiz has to generate questions inside the categories
• update wiki

week 7:

• BUFFER
• final presentation

week 8:

• BUFFER?

Concept 2: Extending Existing Software

Idea

Tutoring system for in the classroom which will be able to come up with questions in a smart way. These questions will be on the average knowledge level of the whole class. The system gives feedback to the teacher about the students enabling him/her to know what students are above average and what students are below. This way the children that are above average can move on in the curriculum independently, while the average children can learn the simple maths with the program and the teacher can focus on helping the children that are not as good. By doing this, the teacher can work more efficiently and more children can be educated at their personal levels. We have found an open source program that resembles Kahoot, and we want to build on this so that eventually it will become an AI program that can generate questions on its own and knows which questions to ask the class. If it is possible (taking into account time and materials) we would like to test this program at an elementary school in Oisterwijk by letting the children make the quiz and check whether or not the program can cluster the children correctly. We can check this by asking the teacher if it is like their expectations. When doing this we can also ask the children and the teacher about their opinion of the program. Some letters will have to be send to the school and/or parents (informed consent?) which we would like to do this week so we can test in week 6. All in all we will test whether or not blended learning can work because we want teacher to be able to work more efficiently as there is a big shortage of elementary teachers.

Persoonlijke planning

week 4

• Abby
  • Figure out how to work with Toohak.
  • Start programming the clustering of the students.
  • Put progress on wiki.
• Christine
  • Figure out how to work with Toohak.
  • Start changing Toohak so that questions can be generated.
  • Put progress on wiki.
• Dennis
  • Figure out how to work with Toohak.
  • Start changing Toohak so that questions can be generated.
  • Progress: Questions can now be generated and input is changed from multiple choice to a textfield of which the numbers are parsed.
• Ellen
  • Make sure there is a correct informed consent for our study.
  • Write a letter to the school and all parents.
  • Put progress on wiki.
• Sophie
  • Send out letters to school and parents (by giving them to a teacher intern).
  • Ask whether we can do the study in a couple of weeks (preferably week 6).
  • Ask what level the class is on.
  • Put progress on wiki.

week 5

• Abby
  • Finish programming the clustering of students.
  • Improve quiz design.
  • Put progress on wiki.
• Christine
  • Finish programming the question generator.
  • Translate all visible text to Dutch.
  • Put progress on wiki.
• Dennis
  • Finish programming the clustering of students.
  • Improve quiz design.
  • Put progress on wiki.
• Ellen
  • Start writing the study report.
  • Prepare for test at school.
  • Put progress on wiki.
• Sophie
  • Prepare for test at school.
  • Improve quiz design.
  • Put progress on wiki.

week 6

• Abby
  • Test in school.
  • Put progress on wiki.
• Christine
  • Take minutes in school.
  • Put progress on wiki.
• Dennis
  • Test in school.
  • Put progress on wiki.
• Ellen
  • Take minutes in school.
  • Put progress on wiki.
• Sophie
  • Test in school.
  • Put progress on wiki.

week 7 who will do the presentation?

• Abby
  • (AI programming).
  • Put progress on wiki.
• Christine
  • Finish study report.
  • Make the presentation.
  • Put progress on wiki.
• Dennis
  • (AI programming).
  • Put progress on wiki.
• Ellen
  • Finish study report.
  • Put progress on wiki.
  • Make sure wiki is clear and contains everything.
• Sophie
  • (AI programming).
  • Put progress on wiki.

week 8

• Abby
  • BUFFER
• Christine
  • BUFFER
• Dennis
  • BUFFER
• Ellen
  • BUFFER
• Sophie
  • BUFFER