Project Description


The main scope of the KOD project is to make a contribution to the thematic area of "Open Platforms and Tools for Personalised Learning" by the design, development and validation of a novel dynamic and adaptive Learning Environment which enables the individual learner to acquire knowledge just-in-time, anytime and anywhere (through the World Wide Web) tailored to his/her personal learning needs.

This is proposed to be achieved by the combination of the knowledge routes concept and software agent technology leading to the establishment of a learner-centred learning system (the KOD System) which

The KOD approach is centred on the extension of current educational content interchange formats (e.g. the Content Packaging Specification currently under extension by international e-learning bodies such as the IEEE LTSC, CEN/ISSS LTW, IMS and ADL SCORM) to facilitate the definition of adaptive and conditional navigation rules taking into account user characteristics (user profiles, testing sessions, etc). These navigation rules will be described together with the learning assets within a single content packaging format, called the Knowledge Packaging (KP) format. The aim of such a notation is to enable instructional designers to describe (in a common, reusable, interoperable and machine readable way) navigational logics which define how knowledge packages (i.e. educational courses, sets of learning assets, etc) can be disaggregated and presented in a different way according to different learner characteristics. As a result, e-learning applications and services can provide different knowledge routes to each individual learner, according to his/her characteristics and needs. The KP format will be used as the interchange format between the specific architectural modules to be developed during the project, which address the needs of all envisaged players of the e-learning arena, namely learners and educational content and/or applications and services providers.


The key objectives of the KOD project are the following

Objective 1. To design, develop and test the KOD learning environment, as a dynamic and adaptable on-line environment which allows the individual learner to acquire knowledge according to his/her personal learning needs.

Objective 2. To provide the methodological framework to support both front-end users (learners) and back-end users (educational content authors, publishers, e-learning service providers, etc), and their organisations, in achieving their educational and training objectives through the use of KOD technology, in a cost-effective and efficient way.

Objective 3. To provide reusable learning material of European added-value in order to support the pilot implementation and the demonstration of the KOD System.

Objective 4. To provide evidence for issues such as efficiency, scalability and cost-effectiveness from the pilot implementation and use of the KOD system in two different learning settings with two different discipline subjects.

Objective 5. To contribute to the emerging European and international standardisation efforts in the area, to interact with other EU initiatives towards the sharing of best practice examples and methodologies at a European level, and to demonstrate the achievements of the project to the public.

Objective 6. To investigate possible opportunities for market exploitation and commercialisation of the KOD Project Results.

Expected Results


KOD Business Model

The KOD Business Model actors are:

The key elements of the KOD underlying business model are depicted in the following figure.

KOD Business Model

As it is shown in the above figure, the KOD business model highlights a process with three main roles and two main control flows, while delivering learning material to learners by means of effective and efficient learning experiences:

KOD Architecture

The conceived KOD architecture addresses the KOD business model with the following objectives:

The architecture of the KOD system aims to address the needs of, and provide tools and methods for regulating services among the 3 players of the KOD business model shown. The rationale of the KOD system is that of providing a resource discovery service dedicated to knowledge brokerage (role B), for matching third party content providers offer (role A) with potential LMS users demand (e-learning managers, role B). Such a service will be based on common content interchange standards, namely the Content Packaging Specification, as well as its foreseen extension to address adaptive educational content definition.

The KOD system layout is represented by the joining of 3 main sub architectures, which can be located in 3 different sites and run by 3 different market players.

A) The Publishing Sub Architecture (Layer A), dedicated to educational content providers willing to publish content in an easy, reusable and interchangeable format to the finest grain necessary for personalisation, reusability and "re-sellability". Layer A will support the Publishing Service players (i.e. content owners, authors, publishers, etc) in describing and packaging together (through the Knowledge Packaging notation) their learning assets metadata, and building different navigational paths. All learning materials described through the Knowledge Packaging notation will be published in a publicly accessible (and monitorable) web space named KOD Publishing Pool (KPP). The Layer A sub-architecture will be centred around a client-server three tier architecture named KOD Packager, enabling users to package their contents in the KOD packaged interchange format (KOD PIF). Each KPP area of each subscribing publisher will host several collections of KOD PIF files describing the publisher's content in terms of building raw assets, their organization and availability.

B) The Brokerage Sub Architecture (Layer B), dedicated to e-learning application service providers (KBS, KOD Brokerage Service Provider) providing the Publishing Services with (i) a dedicated extended version of the KOD Packager, enabling them to test compliance of third party packages and to possibly disassemble and repackage contents in the agreed PIF format, (ii) a set of twin agents (one resident at the Publishing Pool side, the KPP agent, and one at the broker side, the KBR agent) able to shake hands and monitor publication of new PIFs in the KPP area, and (iii) a digital repository (the KOD Broker Repository) managed by the KBR agent and by brokers using a version of the KOD Packager to keep track of all courses/assets available at remote KPP sites. The central KBR repository will be accessible, via web, by subscribing role A publishers willing to access third party content for their publishing purposes, and by role C e-learning service providers willing to enlarge their content offer via the KOD broker functionality. The brokerage service will also be equipped with an environment supporting human brokers to configure agents (KOD ACT, Agency Configuration Toolkit) enabling the user friendly set-up and management of the distributed agents network.

C) The e-learning Sub Architecture (Layer C), supporting role C, i.e. e-learning service providers using "traditional" e-learning platforms as the underlying enabling technology (LMSs). The e-learning Service will be equipped with a User Profiling agent (the KOD UMP, User Metadata profiling Agent) able profile the user (learner), initially and on an ongoing basis. The UMP is capable of inferring the user profile at run time and shake hands with a Knowledge Packages Finder Agent (KPF agent) interfacing with the central KBR broker repository, and able to retrieve the best matching educational content packages available in the central repository, or possibly repackaging on the fly new content layouts compatible with the KOD PIF file. This is then passed to the LMS via the Knowledge Package Delivery (KPD) agent, which assembles actual resources where necessary from the original locations.

KOD System Architecture

Usage Scenarios

Upon subscription, third party content owners (role A) are provided with a KOD Packager, and a Knowledge Packaging Pool Agent (KPPA). The KOD Packager is operating at the content provider site, and supports them to create and store Content Packages in the Knowledge Packaging format, i.e. the IMS CP format and its possible extensions to include adaptive content (i.e. the project will consider the work under progress on adaptive notation, such as Play Rules, the CLEO project in the USA, and the Educational Mark-up Language). The integration of the KOD Packager with existing repositories of assets with internal metadata will also be considered, either for firing queries to the existing internal repositories, or for the batch import of metadata in the KOD Packager. By means of the KOD Packager, tutors will be able to retrieve and assembly existing raw assets (i.e. images, videos, texts, course, lessons) taken from the internal file system into flat navigational hierarchies as those supported in current CP 1.0 (e.g. simple Table of Contents, TOCs), or into conditioned and branched navigation enabled by the Knowledge Packaging format. By the use of the KOD Packager, content owners will be able to publish in explicit forms in a web location (the Knowledge Packages Pools - KPPs) KP manifest files, including descriptions and references to their resources. While the KOD Packager DTD will support the inclusion of resources and other packages with different metadata, a common metadata (e.g. Dublin Core) will be used for the manifest own metadata. Metadata on the commercial availability and reusability of the resources will also be included into the KOD Packager tagging system. The KPPA, on the other hand, will be able to monitor the KPP and return inputs to its colleague agents on the broker side on a regular basis (e.g. upon new entry). This agent could eventually perform all the adaptation and conversion processes before passing a personalised knowledge package to the agent on the broker side (e.g. Manifest metadata conversion to single format if multi formats are considered at the KPP level), for matching specific learner profiles.

The Knowledge Broker (Role B) is provided with an agent (the Knowledge Broker Agent, KBA) able to shake hands and control the network of KPPAs around the net in order to provide input to an internal repository (the Knowledge Brokerage Repository, KBR). The KBA can "unpackage" all KP files arriving from the KPPA network and store all labelled assets in the repository, both at the packaged level and nested sublevels. The KBA normalises all metadata to the internally adopted schema, also recording the original full native XML description. The Knowledge Broker is also equipped with a client-server version of the KOD Packager enabling users (Broker publishers) to access packages in the KBR and re-assembly them into new layouts to be incrementally stored in the KBR preserving original assets metadata. The Consortium is also considering of equipping the Brokerage service with a conformance test bed for testing the conformance of packages. The KB authors (publishers) will be able to access and query the KBR database through Boolean combinations of its internal relational fields, also having an XML viewer of the original metadata available upon for additional filtering. Also, the Knowledge Broker features: (i) a dedicated environment for the configuration and dynamic set up from the broker side of all KOD agents resident on the content provision, brokerage and delivery side (ACT, Agency Configurator Toolkit); (ii) a user profiling service based on a set of twin agents, one, the User Metadata Profile UMP agent, able to accommodate newcomers, profile them and create a IMS complaint learner profile, in turn to be passed to the Knowledge Packages Finder (KPF) agent on the broker side, able to retrieve Knowledge PIFs more suitable to the profiled user and eventually create new packages descriptors on the fly (to be stored in the KBR) and pass them to an agent (the Knowledge Packages Delivery Agent, KPD) on the e-learning service side for user delivery. Subscribing publishers, using the KOD Packager at their premises, will be able to access the central KBR broker repository for retrieving additional content enabled by third party publishers to be merged into their knowledge packages.

Finally, Role C (the e-learning portal manager using a standard LMS compliant to IMS specifications) is equipped with an adapter (KP Importer/adapter) able to receive the KP from the KPD agent and deliver them at run time to the end user (learner). This module, internal to the specific LMS, will be able to import standard Content Packages and possible KOD extensions and build the delivery interface for the profiled user accordingly.

Main Phases

The project workplan adopts an iterative approach, which is driven by user feedback, i.e. it is compliant with the modern user-centred design techniques. The project shall be implemented with the following work packages:

WP1 - KOD System Architecture and Specifications - This work package involves: (i) the 1st phase where user information is fed into the system design (through the initial user needs analysis phase) towards the definition of the abstract KOD model (to be refined through the initial user testing phase of the visual KOD prototype); the information is mainly derived from an analysis of the "KOD context", i.e. the technological, educational, market, etc, aspects of the KOD system; (ii) the definition of the KOD system architecture, and the functional and technical specifications of the KOD System. Specification work focuses on the definition of software tools as common functional, re-usable, interoperable and platform-independent components and building blocks.

WP2 - KOD System Development - This work package involves the development of all software tools which are specified in WP1.

WP3 - Methodology - This Workpackage elaborates on the "Knowledge-on-Demand" methodology for the implementation of the KOD Learning Environment in real learning conditions. Towards this end, special attention shall be given to conducting a survey on the theoretical foundations of personalised learning. This shall also lead to a Handbook of Guidelines for Users (content creators, editors, course builders and instructors), including the background for the definition of Implementation Scenarios and Best Practice Examples.

WP4 - KOD Demonstrator and Pilot Implementation - This WP involves the development of the demonstration and pilot implementation plan, as well as the two KOD demonstrators on telemedicine and knowledge management, based on the tools developed in WP2.

WP5 - Assessment and Evaluation - Involves the assessment and evaluation of the project results. That is, the selection of the specific methods and techniques for the evaluation phase, the development of evaluation scenarios, collection of evaluation results, and finally the analysis of the results and the elaboration of a set of recommendations for the future improvement of the system.

WP6 - Exploitation and dissemination - This work package involves: (i) A set of dissemination activities for the promotion of the KOD project results with emphasis given to the participation and contribution to International Conferences and Exhibitions, Concertation Meetings and clustering activities, and to Standardisation Efforts (CEN/ISSS LTW, IEEE LTSC, IMS, etc) and the PROMETHEUS Initiative, as well as, (ii) Actions towards the investigation of possible market opportunities and commercialisation of the KOD project results.

WP7 - Project Management - This work package involves the Scientific, Technical and Administrative Co-ordination of the KOD Project.