Ideals and an overview of IESS

Jinshan Wu

The Institute of Educational System Science (IESS)

School of Systems Science

Beijing Normal University

Purpose of this presentation

  • To introduce to potential collaborators the vision, key goals and research directions of IESS
    • a new model of education
    • a new model of science of science research and scholarly publications
  • To collect your feedback and critiques

Vision of IESS

  • Helping teachers to teach better and helping students to learn better
  • Helping researchers and R&D administrators to carry out better research and research management
  • Helping enterprises do better management and innovation
  • Please ask me questions at any time

Core concepts

  • The human knowledge highway
  • Levels of knowledge and advanced knowledge generators
    • Level 1 - Factual and procedural knowledge
    • Level 2 - Discipline-specific concepts
    • Level 3 - Research discipline “big picture”
    • Level 4 - General ways of thinking, methods of teaching and learning

Core concepts

  • Top-down connections: higher level knowledge generates and can be extracted/abstracted from lower levels
  • Left-right connections: Pieces of knowledge at the same level are also locally interdependent

Core concepts

  • Meaningful learning: making use of top-down and left-right connections in learning
  • Rote learning: Learning knowledge by repeated practice and memorization, without making use of the above connections
  • See through connections to find the whole, that makes sense to you

About “ability” and “knowledge”

  • In teaching and learning, “ability” refers to being able to make use of knowledge to solve problems, in which one can even create new knowledge as a result
  • If we ask “which ways of thinking are the basis of certain abilities
  • We may identify the core ways of thinking and skills behind the “ability”

About “ability” and “knowledge”

  • “Ability” is the connection between problems and learned knowledge and skills
    • Ways of thinking - 3rd and 4th level knowledge
    • Facts and processes - 1st and 2nd level knowledge
    • As well as the willingness or propensity to address challenging questions

About “ability” and “knowledge”

  • Thus, in our terms, a “ability” is linked directly to knowledge, especially the more advanced knowledge generators, plus the willingness and propensity
  • This is also the spirit of concept mapping: explicitly identifying connections and applying linking phrases to them

Example

AbilityIsKnowledgeGenerator

  • Besides memorizing the facts, this question can be tackled via the concept of “the environment influences human behavior”
  • Thus, the ability to answer this question ultimately becomes a way of thinking, a piece of knowledge

The disciplinary big picture

  • Viewed through the concept of an advanced knowledge generator
    • Typical research subjects
    • Typical research questions
    • Typical methods of analysis
    • Typical ways of thinking
    • How the discipline serves the world and other disciplines

The disciplinary big picture

  • Not everything needs to be learned to establish this big picture of a research discipline
  • Once established, the big picture can be used in learning, using and creating knowledge
  • What are the possible paths to attain this big picture?
  • Inertial knowledge: knowledge that is isolated, not connected to the big picture

A new model of education

  • Meaningful learning targeting advanced knowledge generators over the human knowledge highway
  • Developing students’ own knowledge highway via generative learning
  • Courses and disciplines are structures that emerge from the human knowledge network
  • Every student designs what to become

The final output

  • The human knowledge highway
  • Learning materials attached to concepts and links between them
  • Sequences of learning designed/calculated from a combination of algorithms and experts, may personalized
  • Diagnostic tests designed/calculated from a combination of algorithms and experts, personalized and adaptive

The final output

  • At the level of each piece of knowledge, learning is meaningful and generative, via their connections
  • Make use of top-down and left-right connections in learning
  • Learning is aimed towards the understanding big picture, which can then be readily and widely transferred
  • The Lynkage platform for teaching and learning

Why we need this new model

  • Leaning in order to create and creatively use knowledge
  • Learning in order to appreciate the creation and creative usage of knowledge
  • Teaching students to improve the way they learn, the knowledge foundations they build and their drive to learning

Why we need this new model

  • Repeated usage of knowledge is being replaced by algorithms
  • The challenges of our era call for new creative ways to identify and solve problems
  • Current education often produces next generations who simply recites formulae and conclusions, leaving them only with inertial knowledge

Vision and Mission of IESS

  • Helping teachers to teach better and helping students to learn better
  • Use this new model as the basis for both research and practice
  • See through connections to find the whole

How to implement the new model

  • Task and algorithms for building the highway
  • Develop algorithms to determine sequences of learning and adaptive diagnostic tests
  • Research on meaningful learning, from both behavioral and neuro science

How to implement the new model

  • Promote the underlying concepts as well as the developed system, in and out of schools
  • Extend the systemic approach from learning and teaching to institutional studies of education

Task and algorithms for building the highway

  • Currently manual development, in terms of concept map and wiki

Task and algorithms for building the highway

  • Seek help in developing algorithms to construct the highway from information in textbooks and research papers
  • Annotate the four levels of knowledge
  • Link exercise questions and projects to the highway

The highway as a mathematics model

  • A model describing all the main elements and their relations
    • knowledge network
    • students, teachers
    • questions derived from the practice of teaching and learning

The highway as a mathematics model

  • A platform to develop algorithms to answer the questions for teaching and learning
  • To test out the answers and revise the models to answer the questions better
  • Note: a model firstly represents the data (objects and relations) and then provides a platform to ask and answer questions

Example, Network of Chinese characters

MultilayerFramework

Example, Network of Chinese characters

  • Optimal learning sequence, even personalized
  • Adaptive diagnostic testing system
  • Helps determine what to learn

Learning sequence

  • From the network, we know $a^{i}_{j}=1,0$ when $i$ is/is not a component of $j$
  • Normalize each column to get $\tilde{A}$
  • Solve the following to get $\tilde{W}$, the learning sequence while $W$ is the known frequency of usage \begin{equation} \tilde{W}= \left(1-\tilde{A}\right)^{-1}W= W + \tilde{A}W+\tilde{A}^{2}W+\tilde{A}^{3}W + \cdots \end{equation}
  • Effectively, this calculation considers usage frequency, hierarchical structure and network degree

Learning orders

  • Total number of characters and total usage frequencies

TotalCharactersTotalFrequencies

Experiments on meaningful learning

  • For specific advanced knowledge generators (analogy, if-then, Chinese characters), comparing meaningful and rote learning
    • Learning cost (emotional and cognitive burden)
    • Learning results (transfer learning, transfer creation, grades)
    • Characteristic brain activity patterns in learning, using, and creating knowledge

Experiments on meaningful learning

  • For a set of knowledge, such as a group of Chinese characters, experiments on meaningful learning and sequences of learning from algorithms
  • For a subject in certain grade, such as elementary school mathematics
  • Experimental validation of the diagnostic testing algorithm
  • Modules of the Lynkage platform and experiments

Experiments on meaningful learning

  • Misconception and its resolving or suppression
  • From knowing to applying - beyond simple repetition
  • Multi-brain synchronization, combining teaching and learning process in meaningful learning

Extended to institutional studies of education

  • Manual annotation of the knowledge levels of teaching activities - also an AI-supported system
  • How teacher allocate their time?
  • What is ultimately learned from schools?
  • Teacher’s education and school assessment that promote meaningful learning

The “New” and “Model”

  • A mathematical model for all the major elements and their relations in teaching and learning
  • Turns questions on teaching and learning into math problems
  • Find ways to solve the math problems using the model
  • Experimentally test the solutions and the methods of analysis
  • Push teaching and learning to be more scientific and systematic, using modeling and testing
  • Apply the same model to research and practice
  • Please stay tuned for “Lynkage: meaningful learning on the human knowledge highway”

The Three-Layer Network for Scientometrics

MultilayerFramework

The Framework: Layers

  • The concept layer: concepts with their logical connections, such as theorems connected by ‘prove’
  • The paper layer: papers connected by citations
  • The author layer: researchers connected by mentor-mentee relationships
  • Inter layer connections: author-writes-paper, paper-studies-concepts

The Framework: What can be done

  • Clustering papers and concepts into topics/fields/disciplines
  • Evaluate creativity of papers/authors
  • Get an overview of the whole field, for researchers and administrators
  • Ultimately help the development of science
  • Can also help teachers and students

The Framework: why a model

  • A mathematical model for the major elements of the Science of Science
  • Scientometric questions become math problems in terms of this mathematical model
  • Solve, test and generalize them
  • A math language that describes the data, questions, method of analysis and ways of thinking

Examples: several systems

  • Math theorems, connected by ‘prove’, with papers and authors
  • Connections between chemical reactions and reactants, with papers and authors
  • Networks of connections between diseases and treatments/medicines, with papers and authors

Examples: method of analysis

  • Collect data to build up the underlying network
  • Ask questions and represent them in terms of the network
  • Seek ways to solve the questions
  • Test, and systemize if necessary
  • Network analysis: A combination of direct and indirect connections

The “New” and “Model”

  • A mathematical model for all the major elements and their relations in scientometrics
  • Turns scientometric questions into math problems
  • Find ways to solve the math problems using the model
  • Experimentally test the solutions and the methods of analysis
  • Push scientometrics to be more scientific and systematic, using modeling and testing
  • Apply the same model to research and practice
  • Please stay tuned for “Lynkage for Sci$^{2}$”

A new model of publications

  • Papers published together with concept maps as abstracts
    • Show which concepts are being studied
    • Show conclusions and how they are supported
    • Easily integrated into research discipline concept network
  • Can be used for books and other sources

Research paper database

  • ReseaCmap, similar to chemical reaction database, Reaxys, SciFinder
    • Each paper takes several concepts in as input and generates some concepts as output
    • Far more accurate and informative than keywords
    • Domain researchers can use it to help their research
    • Also useful to scientometric researchers and R&D administrators
    • Even a robot researcher?

IESS birdview

AnOverviewOfIESS

Summary of the new models

  • Mathematically modeling the main elements and their relations in education and sci$^2$
  • Turning the relevant questions into math problems
  • Solving, testing and systemizing them, thus, making them more scientific
  • The same framework can potentially be used to serve enterprises

Questions

  • Thank you for your time and feedback
  • Take-home msg: together, we can make a difference to education and sci$^{2}$
  • Helping teachers/students/researchers to teach/learn/research better
  • Making the world a better place
  • For more information, Big Physics and IESS