Learning objects and metadata framework - Mohammed Kharma
Download as DOCX, PDF0 likes868 views
The document discusses learning objects, metadata frameworks, and automatic metadata generation. It defines learning objects and explains that metadata helps increase reuse of learning objects by providing contextual information. It then discusses common metadata standards like LOM and Dublin Core, and frameworks that can be used to automatically generate metadata for learning objects based on user interactions and behaviors.
![Learning objects and metadata framework: concept, definition, design and implementation<br />Mohammed Fahmi Kharma<br />University of Al-Quds: Department of Computer Science<br />mkharma@science.alquds.edu, Moh.kharma@gmail.com <br />Abstract<br />Learning Object concept aimed to decrease the cost of developing electronic learning (educational) material by enhancing content re-use form during construction of learning material where for example any learning material may constructed from multimedia content, instructional content, other learning objects, instructional software and tools and persons/organization. So one of the main achievements in learning object technology will be enhancing learning object re-use. The baseline of the enhancement comes from strongly understanding the way of using learning objects and metadata by end users. <br />Introduction<br />E-learning is commonly referred to the intentional use of networked information and communications technology in teaching and learning [1]. E-learning term is referring to not only for online, virtual, distributed networked or web-based learning, but also to any educational activities shown in below diagram related to online or offline individuals and groups working activities that asynchronously/synchronously via networked or standalone computers and other electronic devices. <br />E-learning modalities as in Fig.1 can be one of four modalities, Individualized self-paced e-learning offline applied when an individual learner is using learning resources while there is no Intranet or the Internet connection is being used such as a database or a computer-assisted learning package offline as CD or DVD .Individualized self-paced e-learning online is the second type of E-learning where the individual learner is using learning resources using internet or intranet online such as accessing course content online. Thirdly, Group-based E-learning<br />synchronously same as learners who are using instance messaging or videoconference where the groups of learners are communicating together in real time over the Internet or Intranet. Fourthly, Group-based e-learning asynchronously like on-line discussions viaFig.1 - E-Learning modalities[1]<br />electronic mailing lists where groups of learners are using Internet or Intranet for communication and exchanging but not in real time interaction<br />Learning Objects and Metadata<br />Learning Object is defined as any entity, digital or non-digital, that may be used for learning, education or training (IEEE). Learning Object concept aimed to decrease the cost of developing electronic learning (educational) material by enhancing content re-use form during construction of learning material. One of the main achievements in learning object technology will be enhancing learning object re-use. The baseline of the enhancement comes from strongly understanding the way of using learning objects and metadata by end users.<br />High usability of learning objects components depend on the amount of contextual information that the object have. If we are talking about a story for example, we will find it carried out to targeted users and contains specific context and objectives can’t be used in another context while if we talk about parts or the raw data of the story element like animations, text...etc. it will have less context information compared to story in general. Therefore, this will allow the users to re-use the small components much better than large component since small component will have less context information.<br />Metadata refers to the controlled taxonomy and related vocabulary used to describe learning objects [3] or it’s a descriptive data used to annotate another data so can be used in different context like in content indexing by search engines. The first step in metadata design is to create an application profile by customizing a base schema to fit the needs of target users, and select the metadata elements that will be included in the customized profile to describe stored learning objects considering the element value Domains. As a concrete example, ARIADNE use automatic metadata generation framework to create a metadata for uploaded materials to be used to annotating the uploaded material. <br />Metadata standards in learning object domain<br />Number of metadata standards has been developed in E-learning learning object area since early <br />days of internet; in 2002, IEEE finally approved Learning Object Metadata (LOM) as an international standard for LOM which is wildly accepted and being used in E-learning domain. In addition to Dublin Core metadata standard which includes two levels: Simple and Qualified. Qualified Dublin Core includes three additional elements (Audience, Provenance and RightsHolder) and SimpleFig.2 - Dublin Core schema<br />Dublin Core included fifteen elements (Fig.2). Dublin Core schema consists fifteen terms where each term has designed to express simple textual information about the learning object. <br />IEEE LOM standard has been originated in Joint cooperation work between IMS Global Learning Consortium and European ARIADNE project in 1995. LOM constructed from 76 elements spitted in nine categories, each element data type and content can be specified to have certain syntax and vocabularies (Fig.3).<br />Fig.3 - IEEE LOM Schema.[7]<br /> <br />As a comparison between the above two standards, the terms in Core Dublin are defined to be used without any relation between them. In LOM standard, the meaning of elements depends on the location of it within the structure of the metadata<br />IMS Learning Design and Specifications<br />IMS LD specification aimed to provide a model to define the activities and tasks and structures and associated assignment to roles and the workflow of ‘learning design’ as a unit of learning and can be used to represent heterogeneous approaches to be adaptive in eLearning. adaptive E-learning method aimed to enhance the performance of a pre-defined criteria like educational, economic etc. throughout creating learning experience to the student or tutor depend on configuration of a set of elements in a specific period. Daniel Burgos Et al. have been successfully set up a list of adaptation in E-Learning systems that can adjust the set of rules of dependencies between learners and get the best learning experience and they have presented number of features and definitions for adaptive learning and also they have defined how features and elements of adoptive learning can be addressed by IMS LD.<br />Automatic metadata generation<br />Automatic metadata generation process is one of hot topics in E-learning domain where metadata creation done on automatic base rather than the traditional ways which required massive human interaction and effort. But the automation for the process of metadata generation will make the annotation and tagging of the learning objects more simple and online to have better outputs. Tree basics property should be achieved by such automatic metadata generation systems as following: (a) general: based on widespread e-learning standards; (2) automatic: metadata annotation should not require human intervention; and (3) in-terpretable: metadata should be both human- and machine-readable.[5]<br />Interesting framework for automatic metadata generation has been proposed by S.A. RILEYa Et. al in there paper[4] where they implemented an Intelligent Learning Object Guide (iLOG) to be used as framework for automatic learning object annotation generation based on empirical real-world usage by 200 students to collect these students integration with eight learning objects. Current studies that use standards for tagging LO like IEEE Learning Object Metadata (LOM) prove that such standard is incomplete, inaccurate, and not machine-interpretable. A system that automatically tags LOs with empirical usage metadata should be general based on large spectrum of e-learning standards and automatic and interpretable for human and machine.<br />Fig.4 - The Intelligent Learning Object Guide (iLOG) Framework.[5]<br />S.A. RILEYa Et. al implementation for automatic empirical usage metadata have consist a MetaGen component where it is a processes for log file data in order to generate empirical usage metadata using predictive rule mining and then sent back to the LO wrapper and LO wrapper component for logging student interactions and updating the LO metadata of iLOG as Fig.4 illustrate above.<br />Supporting Tools and Systems for LOM<br />There are many supporting tools for learning object metadata, IMS Editor Vimse (ImseVimse) from the Conzilla project is one of earliest and well known developed editor as LOM editors. Also LOM-Editor is a learning object metadata editor has been developed by Multimedia Communications and was written in Java. In addition, RELOAD Editor which is a combined content packager and metadata editor that can create, import, edit, export LOM and content packages and supports IMS Metadata 1.2.2 XML binding. About metadata generation, The Learning Object Metadata Generator (LOMGen) is the leading tool for automatic harvesting and constructing the LOM from the Web [5].<br />Summary and Conclusion<br />After the continuous investments and achievements in E-learning domains. And the proven success and benefits for it. And as an available environment, we suggest to apply automatic metadata generation for in Moodle, and make Moodle integrations with some of learning resources provider such as ARIADNE and GLOBE where around a million learning resources are available, and import a found resource directly into the LMS. By this, LMS will be interesting environment where large volume of resources will be available automatically for the user and the automatic tagging process also will enhance the search experience of the end users.<br />References<br />E-Learning, A Guidebook of Principles, Procedures and Practices. Som Naidu, The University of Melbourne , 2006.](https://image.slidesharecdn.com/synthesispaper-mohammedkharma-110813161845-phpapp02/85/Learning-objects-and-metadata-framework-Mohammed-Kharma-1-320.jpg)
Learning objects and metadata framework - Mohammed Kharma
- 1. Learning objects and metadata framework: concept, definition, design and implementation<br />Mohammed Fahmi Kharma<br />University of Al-Quds: Department of Computer Science<br />mkharma@science.alquds.edu, Moh.kharma@gmail.com <br />Abstract<br />Learning Object concept aimed to decrease the cost of developing electronic learning (educational) material by enhancing content re-use form during construction of learning material where for example any learning material may constructed from multimedia content, instructional content, other learning objects, instructional software and tools and persons/organization. So one of the main achievements in learning object technology will be enhancing learning object re-use. The baseline of the enhancement comes from strongly understanding the way of using learning objects and metadata by end users. <br />Introduction<br />E-learning is commonly referred to the intentional use of networked information and communications technology in teaching and learning [1]. E-learning term is referring to not only for online, virtual, distributed networked or web-based learning, but also to any educational activities shown in below diagram related to online or offline individuals and groups working activities that asynchronously/synchronously via networked or standalone computers and other electronic devices. <br />E-learning modalities as in Fig.1 can be one of four modalities, Individualized self-paced e-learning offline applied when an individual learner is using learning resources while there is no Intranet or the Internet connection is being used such as a database or a computer-assisted learning package offline as CD or DVD .Individualized self-paced e-learning online is the second type of E-learning where the individual learner is using learning resources using internet or intranet online such as accessing course content online. Thirdly, Group-based E-learning<br />synchronously same as learners who are using instance messaging or videoconference where the groups of learners are communicating together in real time over the Internet or Intranet. Fourthly, Group-based e-learning asynchronously like on-line discussions viaFig.1 - E-Learning modalities[1]<br />electronic mailing lists where groups of learners are using Internet or Intranet for communication and exchanging but not in real time interaction<br />Learning Objects and Metadata<br />Learning Object is defined as any entity, digital or non-digital, that may be used for learning, education or training (IEEE). Learning Object concept aimed to decrease the cost of developing electronic learning (educational) material by enhancing content re-use form during construction of learning material. One of the main achievements in learning object technology will be enhancing learning object re-use. The baseline of the enhancement comes from strongly understanding the way of using learning objects and metadata by end users.<br />High usability of learning objects components depend on the amount of contextual information that the object have. If we are talking about a story for example, we will find it carried out to targeted users and contains specific context and objectives can’t be used in another context while if we talk about parts or the raw data of the story element like animations, text...etc. it will have less context information compared to story in general. Therefore, this will allow the users to re-use the small components much better than large component since small component will have less context information.<br />Metadata refers to the controlled taxonomy and related vocabulary used to describe learning objects [3] or it’s a descriptive data used to annotate another data so can be used in different context like in content indexing by search engines. The first step in metadata design is to create an application profile by customizing a base schema to fit the needs of target users, and select the metadata elements that will be included in the customized profile to describe stored learning objects considering the element value Domains. As a concrete example, ARIADNE use automatic metadata generation framework to create a metadata for uploaded materials to be used to annotating the uploaded material. <br />Metadata standards in learning object domain<br />Number of metadata standards has been developed in E-learning learning object area since early <br />days of internet; in 2002, IEEE finally approved Learning Object Metadata (LOM) as an international standard for LOM which is wildly accepted and being used in E-learning domain. In addition to Dublin Core metadata standard which includes two levels: Simple and Qualified. Qualified Dublin Core includes three additional elements (Audience, Provenance and RightsHolder) and SimpleFig.2 - Dublin Core schema<br />Dublin Core included fifteen elements (Fig.2). Dublin Core schema consists fifteen terms where each term has designed to express simple textual information about the learning object. <br />IEEE LOM standard has been originated in Joint cooperation work between IMS Global Learning Consortium and European ARIADNE project in 1995. LOM constructed from 76 elements spitted in nine categories, each element data type and content can be specified to have certain syntax and vocabularies (Fig.3).<br />Fig.3 - IEEE LOM Schema.[7]<br /> <br />As a comparison between the above two standards, the terms in Core Dublin are defined to be used without any relation between them. In LOM standard, the meaning of elements depends on the location of it within the structure of the metadata<br />IMS Learning Design and Specifications<br />IMS LD specification aimed to provide a model to define the activities and tasks and structures and associated assignment to roles and the workflow of ‘learning design’ as a unit of learning and can be used to represent heterogeneous approaches to be adaptive in eLearning. adaptive E-learning method aimed to enhance the performance of a pre-defined criteria like educational, economic etc. throughout creating learning experience to the student or tutor depend on configuration of a set of elements in a specific period. Daniel Burgos Et al. have been successfully set up a list of adaptation in E-Learning systems that can adjust the set of rules of dependencies between learners and get the best learning experience and they have presented number of features and definitions for adaptive learning and also they have defined how features and elements of adoptive learning can be addressed by IMS LD.<br />Automatic metadata generation<br />Automatic metadata generation process is one of hot topics in E-learning domain where metadata creation done on automatic base rather than the traditional ways which required massive human interaction and effort. But the automation for the process of metadata generation will make the annotation and tagging of the learning objects more simple and online to have better outputs. Tree basics property should be achieved by such automatic metadata generation systems as following: (a) general: based on widespread e-learning standards; (2) automatic: metadata annotation should not require human intervention; and (3) in-terpretable: metadata should be both human- and machine-readable.[5]<br />Interesting framework for automatic metadata generation has been proposed by S.A. RILEYa Et. al in there paper[4] where they implemented an Intelligent Learning Object Guide (iLOG) to be used as framework for automatic learning object annotation generation based on empirical real-world usage by 200 students to collect these students integration with eight learning objects. Current studies that use standards for tagging LO like IEEE Learning Object Metadata (LOM) prove that such standard is incomplete, inaccurate, and not machine-interpretable. A system that automatically tags LOs with empirical usage metadata should be general based on large spectrum of e-learning standards and automatic and interpretable for human and machine.<br />Fig.4 - The Intelligent Learning Object Guide (iLOG) Framework.[5]<br />S.A. RILEYa Et. al implementation for automatic empirical usage metadata have consist a MetaGen component where it is a processes for log file data in order to generate empirical usage metadata using predictive rule mining and then sent back to the LO wrapper and LO wrapper component for logging student interactions and updating the LO metadata of iLOG as Fig.4 illustrate above.<br />Supporting Tools and Systems for LOM<br />There are many supporting tools for learning object metadata, IMS Editor Vimse (ImseVimse) from the Conzilla project is one of earliest and well known developed editor as LOM editors. Also LOM-Editor is a learning object metadata editor has been developed by Multimedia Communications and was written in Java. In addition, RELOAD Editor which is a combined content packager and metadata editor that can create, import, edit, export LOM and content packages and supports IMS Metadata 1.2.2 XML binding. About metadata generation, The Learning Object Metadata Generator (LOMGen) is the leading tool for automatic harvesting and constructing the LOM from the Web [5].<br />Summary and Conclusion<br />After the continuous investments and achievements in E-learning domains. And the proven success and benefits for it. And as an available environment, we suggest to apply automatic metadata generation for in Moodle, and make Moodle integrations with some of learning resources provider such as ARIADNE and GLOBE where around a million learning resources are available, and import a found resource directly into the LMS. By this, LMS will be interesting environment where large volume of resources will be available automatically for the user and the automatic tagging process also will enhance the search experience of the end users.<br />References<br />E-Learning, A Guidebook of Principles, Procedures and Practices. Som Naidu, The University of Melbourne , 2006.
- 2. Representing adaptive eLearning strategies in IMS Learning Design. Daniel Burgos, The Open University of The Netherlands 2009
- 3. DESIGN PRINCIPLES FOR AUTHORING DYNAMIC, REUSABLE LEARNING OBJECTS. Tom Boyle, Learning Technology Research Institute (LTRI)
- 4. Investigating teachers' understanding of IMS Learning Design: Yes they can!?. Michael Derntl Et. al.2010
- 5. Intelligent Learning Object Guide (iLOG): A Framework for Automatic Empirically-Based Metadata Generation. S.A. RILEYa Et. al.
- 6. Tree View Editing Learning Object Metadata. Zeynel Cebeci and Yoldas Erdogan, 2005.
- 7. IMS Meta-data Best Practice Guide for IEEE 1484.12.1-2002 Standard for Learning ObjectMetadata, http://www.imsglobal.org/metadata/mdv1p3pd/imsmd_bestv1p3pd.html