Improving Learning Object Description Mechanisms to Support an Integrated Framework for Ubiquitous Learning Scenarios María Felisa Verdejo Carlos Celorrio.

Improving Learning Object Description Mechanisms to

Support an Integrated Framework for Ubiquitous

Learning Scenarios

María Felisa VerdejoCarlos Celorrio

Emilio Julio Lorenzo

LTCS Group - UNED

WMUTE 2006, Athens

Outline

Introduction Enlace Project Ubiquity in Learning Processes Educational network infrastructure

Metadata for Tool’s Objects Automatic Metadata Generation Integration by Tool Wrapping Conclusions

Introduction

Mobile and ubiquitous technologies have demonstrated potential to enrich learning and teaching inside and outside the classroom.

An increasingly number of experiences are being reported.

However all these technologies are often a range of unintegrated options.

ENLACE Project

Spanish National Research Project http://www.enlace.uned.es

Explores the design of an innovative educational environment

Technology seamlessly supporting the flow of learning activities in different contexts and settings

Enlace Project Students are engaged both in

individual and collaborative activities compiling information analyzing and representing

data building models to explain and

predict behaviour exchanging and discussing

their findings linking their view with real

problems

Ubiquity in Learning Processes The learning workflow includes

activities for a long term period (to be carried out

either in sequence, parallel, or overlapping in time)

in different scenarios (classroom, home, computer lab, field trips…)

Ubiquitous context challenging the integration issue through the school curriculum location time social organization levels across devices interoperability connectivity

Ubiquity in Learning Processes

Smooth flow throughout the scenarios of activities

Users can create artefacts that will be reused in other tools, in other devices, or even different students in different contexts.

Sustain seamless data and tool interoperability, through an educational network infrastructure a variety of mobile devices computer tools connectivity conditions

Educational network infrastructure

The backbone of the technological infrastructure is a Learning Object Repository (LOR)

a) a persistence mechanism for tools

b) a web interface foraccessing artefacts

Metadata is akey factor

Internet (Web Services)

Main LOR

Modeling Space

Field trip

Computer lab

WI-FI ad hoc (http)

PdaCards CARDS Local LOR

Internet

Cards

A

B

Outline

Introduction Metadata for Tool’s Objects Automatic Metadata Generation Integration by Tool Wrapping Conclusions

Metadata for Tool’s Objects

LOR as a service for external tools...

1. a declarative way for defining specific metadata.

2. mechanisms for automatic metadata generation

GOAL: generalized framework for the external tools.

Metadata for Tool’s Objects

A LO type in the repository defines the description for a class of LOs General Metadata Schema + other Tool’s Specific

Metadata schemas the format of content (i.e. multimedia material, XML

files, models following tool specific formats).

Metadata for Tool’s Objects

Metadata items are grouped in Metadata Schemas.

Each metadata item has a value type Text, for storing any kind of string text. Number, for integer or real numeric values, possibly

delimited with maximum and minimum values. Predefined formats of interest, as for instance, date, time

and position for storing GPS data. Vocabulary, for predefined set of values. Terms of an ontology

Metadata for Tool’s Objects

The LOR provides mechanisms for definingnew LO types for tools ina declarative way

An example: Observation Card Object from CARDS tool

Specific Metadata

General Metadata

Outline

Introduction Metadata for Tool’s Objects Automatic Metadata Generation

The Method Extraction sources Conflict resolution criteria

Integration by Tool Wrapping Conclusions

Automatic Metadata Generation

Metadata is seen as a key feature for enhancing the search in digital repositories.

How this metadata is created? 1) The author of the resource creates the metadata using

editing tools 2) A professional cataloguer creates metadata (as

librarians currently do) and 3) Methods for automatic generation of metadata are

applied. We want a completely automatic metadata

generation mechanism

The Method

Automatic metadata generation process includes two steps:

1. Tool extraction of metadata field values.

2. Combining metadata field values.

Extraction sources

LO types in the LOR includes information about how the metadata fields will be populated.

Extraction sources for automatic metadata generation: 1) Tool’s own context 2) Default values 3) The object itself 4) The portal context 5) Inference from an ontology that models the different

concepts underlying the metadata domains..

Conflict resolution criteria

Defined at the moment of the LO type declaration.

A list of ordered candidates.

Tool’s metadata generated values has preference over the LOR’s generated ones.

Outline

Introduction Metadata for Tool’s Objects Automatic Metadata Generation Integration by Tool Wrapping Conclusions

Integration by Tool Wrapping

Many tools are not prepared for interoperating directly with the LOR’s web services interface.

The idea for handling this problem is to use a mechanism allowing 1) to define an “external” metadata model suitable to

describe potential outcomes from that tool 2) to wrap the non-compliant tool to produce

a) appropriate metadata values from its output using the metadata model

b) to generate input data to feed the tool from a selected LO.

Integration by Tool Wrapping

CARDS is an authoring system for defining models for data collection and the metadata associated to these models.

ModellingSpace is a collaborative modelling tool to create models to represent and understand natural phenomena, physics laws, trigonometric rules, etc.

CARDS

MODELING SPACE

Integration by Tool Wrapping

An example of this flow in a topographic profiling activity.

CARDS TOOL

MODELING SPACE TOOL

LOR

PDACARDS

Outline

Introduction Metadata for Tool’s Objects Automatic Metadata Generation Integration by Tool Wrapping Conclusions

Conclusions

Storing and retrieving the products generated by students from different devices and tools

Many diverse tools can be potentially integrated in the networking infrastructure, either by web services interoperability or by tool wrapping.

The definition of an LO type allows to specify and implement a variety of automatic metadata generation methods to create a rich description of an artefact.

Conclusions

Perspectives Scenario and activity design Pilot evaluation Broaden participation Tool integration Agent technology

Configuration layer with functionality adaptation to dynamically deploy the required services customized for a certain learning scenario

Improving Learning Object Description Mechanisms to

Support an Integrated Framework for Ubiquitous

Learning Scenarios

María Felisa VerdejoCarlos Celorrio

Emilio Julio Lorenzo

LTCS Group - UNED

WMUTE 2006, Athens