semantic learning instructor: professor cercone razieh niazi
TRANSCRIPT
Outline
Introduction Issues in the Current State of Knowledge
Discovery Intellectual Knowledge Discovery Learning Objects Granularity Issue Proposed Solution
Problems in the current state
Knowledge discovery: difficult Information overload:
Most of the problems which we have finding a path though the huge amount of information currently available not only on the Web but in books, newspapers, television and films.
Evaluating these information needs skills
Information authentication Indexing facilities are used in conventional systems like
libraries, and in search engines. sites exist which present themselves as impartial research conduits when in fact they are funded by commercial and other interests.
Knowledge neighborhoods Customization of information discovery:
Given the amount of information available, the problem of matching learner to material, which is relevant to his or her needs at a particular point in time, becomes more and more required.
My Model: Intelligent Learning Environment
Interconnecting Knowledge
Neighborhoods
Interconnecting Knowledge
NeighborhoodsAutomatic
Learning Object Aggregation
Automatic Learning Object
Aggregation
PersonalizationPersonalization
AdaptabilityKnowledge Navigator
Knowledge Navigator
Collective IntelligenceCollective
Intelligence
ContextualizationContextualization
Intelligent Learning
Environment
Knowledge GenerationKnowledge Generation
E-learning Platforms
M-learning Platforms Pervasive-learning Platforms
Learners(Human) Devices Agents
A Web of Knowledg
e
A Web of Knowledg
e
Semantic learning Platforms
Learners:
Dream comes True!!!
Basic components: Annotated educational resources, a means of reasoning about these, and a range of associated services.
The basic step is having ability to aggregate learning object.
Learning Objects
What is a Learning Object? small units of learning resources self-contained are reused are aggregated, and combined
Reusable Learning Object
"reuse" means placing a learning object in a context other than that for which it was designed
What “Reusable Learning Object” brings for us? Personalized Learning Customized Lessons Interconnecting Knowledge Neighborhoods Generate Knowldge
Current State of Learning Object
Learning objects are identified with metadata so that they can be referenced and searched both by authors and learners.
Cisco Model
Scorm
SCORM stands for Sharable Content Object Reference Model, initiated by Advanced Distributed Learning (ADL) specification group.
Issues: the current design of SCORM has resulted in:
• the slow pace• high cost developing of learning objects• not able to be tailored to individual needs
LOM
LOM: IEEE Learning Object Metadata Learning Object Metadata is a data model encoded in
XML and used to describe learning objects. Developed by IEEE supports reusability of learning
objects, aids discoverability and facilitates interoperability in the context of online learning management systems
Issues with the Current State
A concept can be described by two dimensions including:
Intention: Set of concept’s attribute and values
Extention: A set of objects that belongs to the concept
The current metadata standards provide the extension of the objects.
LO are considered as a lecture or media,… They can not aggregate to make a personalized
lesson Indeed, the major issue is:
Granularity !!
In the philosophical perspective: Granular computing attempts to extract and formalize
human thinking.
In the methodological perspective: It concerns structured problem solving.
In the computational perspective: It is a paradigm of structured information processing.
It addresses the problems of information processing in the abstract
Granular computing exploits structures in terms of granules, levels, and hierarchies based on multilevel and multi-view representations
A granule normally consists of elements that are drawn together by indistinguishability, similarity or functionality
Writing may be viewed as a problem solving process and task.
A simple idea is described by a paragraph consisting of several sentences.
A point-of-view is jointly described and supported by several ideas.
Tasks
Building Granular learning objects: Annotation Metadata based on standards i.e: IMS 1st level Granulation Feature Extraction Functional Representation of Granules Hierarchical Structure Of Granules Description language for Learning Objects
Publish Universal Repository for published learning objects
Discovery Learning Path 2nd level granulation (Rough-based approach)
LORDLORD
LOLOLearner
PublishDiscovery
Retrieve
LODL
Learning Path
Proposed Model- Reusable Learning Objects
Text
GranulateAnnotate
Feature Extraction Functional Representation of granules
Design TimeRun Time
Publish
Build HierarchicalStructure Of Granules
Publish LODL (Learning Object Description
Language)
Metadata on Text
Metadata on Text
LORD(Learning Object Repository and Directory
Build
Discovery
Rough set Granulation
Learning Path
Customized Lesson
Proposed Model: Functional Representation of the Learning Objects
Endpoint: https://wiki.cse.yorku.ca/course_archive/2010-11/W/4403/lectures
Endpint: http://www.fuzzy-logic.com/Ch1.htm
http://www.cs.cmu.edu/Groups/AI/html/faqs/ai/fuzzy/part1/faq-doc-2.html