ALOA

Our paper "ALOA: A Web Services Driven Framework for Automatic Learning Object Annotation" has been published by Springer (.pdf, project homepage).

Abstract:

Generating learning object metadata is a complex task to be done manually. An automated approach is therefore required. This work presents the concepts and ideas behind the automatic generation of metadata. We propose a Web Services driven framework for IEEE LOM-compliant automatic learning object annotation called ALOA. The primary focus has been on the flexibility and extensibility of the framework, such that new metadata generation services can easily be plugged into the basic system.

ECTEL and MUPPLE 2008

Last week, I attented the ECTEL conference in Maastricht, the Netherlands. I presented there our paper "ALOA: A Web Services Driven Framework for Automatic Learning Object Annotation" (slides are available on Slideshare, ALOA can be tested here) within a Firehose session. Here how it works, as described by the conference organizers:

The FireHose format aims to make sessions more interactive than standard paper sessions. There are 4 to 5 presentations in each FireHose session and the session last for 90 minutes. The presenters prepare a poster and/or a demo on their laptop and/or any kind of material that supports their presentations as well as a 5 minutes introductory talk that they give to the session chair just before the session with an USB stick. The session chair ask each presenter to give a 5 minutes talk with the beamer. It's really 5 minutes: the goals, the method and one example of result. Just enough for participants to decide if they want to know more. When the introductory round is finished, the presenters move to their poster/table and attendees join one of the presenters for a 15 minutes period for a longer presentation and small group discussions. After 15 minutes, the session chair shouts "ROTATE" and attendees move to another table in the room.

I'm not sure if this format has been applied elsewhere, but it was indeed an excellent idea. From my experience, it's much better that the normal presentation format where you talk for some time and then you just get a couple of questions from the audience. The FireHose format enables much more interaction and lively discussions with almost everyone in the audience.

I also attended the MUPPLE workshop which was indeed a very successful event, very well organized by my dear friends Fridolin Wild and Matthias Palmer. I met great people there working on PLE issues. Among others, I much enjoyed the talks by Tony Hirst who talked about flexible learning environments and used my PLE diagram as an example in his slides, Felix Mödritscher who introduced LISL as a new scripting language to define activities within a PLE, Hannes Ebner who talked about metadata management, Scott Wilson who talked about widget integration and Graham Attwell who talked about mashup PLEs. I know Scott and Graham from their blogs, and it was a great pleasure for me to finally meet them in person there. In case you're wondering whether Scott and Graham are as good as the image that you can build of them from the readings of their blogs and scientific publications, here's the answer: NO, they're better. I learned many things from them during the workshop. We had in the end a great discussion round about challenges and future directions in PLE research. I talked about PLEF; a conceptual framework that can support learners in taking control over their learning activities by creating and managing their own PLEs. Will write more about this framework in a later post.

ALOA: A Web Services Driven Framework for Automatic Learning Object Annotation

View SlideShare presentation or Upload your own. (tags: metadata)

Google Translate and ALOA

Google Operating System is reporting that Google has switched the translation system on its Google Translate service to its own technology. Previously, the site used Systran for almost all of its translation processing, except Arabic, Chinese, and Russian.
Our ALOA system, the Web Services driven framework for IEEE LOM compliant automatic matadata generation, is using Google Translate for its translation service. Here I use the ALOA framework to generate the LOM summary metadata from the MIT course "Database Systems" in 5 different languages; English, French, German, Arabic, and Spanish. The result can be seen in the screenshot below. I´m very familiar with these languages and I can say that the result is quite good for a machine translation.

ALOA Framework

I´ve got several requests on how to use the ALOA Web Services. ALOA is the Web Services driven framework for IEEE LOM compliant automatic matadata generation. More information about this framework is availabe here. A detailed description of the ALOA Web Services can be found here. I´ve also written a simple class that uses the ALOA Web Services to generate the LOM summary metadata from Stephen Downes´famous article E-Learning 2.0 in 3 different languages; French, German, and English. You can see the result below or test it yourself at the ALOA project homepage.

Code:

import java.rmi.RemoteException;

import i5.prolearn.aloa.samgiservice.SamgIPortProxy;
import i5.prolearn.aloa.samgiservice.types.MetadatasourceId;
import i5.prolearn.aloa.samgiservice.types.URLMetadatasourceId;
import i5.prolearn.aloa.samgiservice.types.samgiFaults.SamgiFault;


public class ALOAWS {

public static void main(String[] args) throws RemoteException, SamgiFault {

URLMetadatasourceId metadatasourceId =
new URLMetadatasourceId();
metadatasourceId.setMetadatasourceIdName("test");
metadatasourceId.setLoURLLocation(
"http://www.downes.ca/post/31741");
SamgIPortProxy sampleSamgIPortProxyid = new SamgIPortProxy();
sampleSamgIPortProxyid.setEndpoint(
"http://eiche.informatik.rwth-aachen.de:3333/ALOA/services/SamgIPort");
String[] targets = new String[] {"english", "german", "french"};

sampleSamgIPortProxyid.setLanguages("1234", targets);

String[] types = new String[] {"SUMMARY"};
String res = sampleSamgIPortProxyid.getMetadata(
"1234", new MetadatasourceId[] {
metadatasourceId }, types);
System.out.print(res);

}

}

Result in XML:

ALOA: A Web Services Driven Framework for Automatic Learning Object Annotation

Yesterday, Nanda Firdausi Muhammad successfully defended his master thesis "ALOA: A Web Services Driven Framework for Automatic Learning Object Annotation". Under my supervision, Nanda has designed, implemented, and evaluated a Web Services driven framework for IEEE LOM compliant automatic matadata generation. The primary focus has been on the flexibility and extensibility of the framework, such that new metatata generation services/modules can easily be plugged into the basic system.

The proposed solution and the implemented system fulfill these requirements; the system already implements different modules and is capable of generating a big part of the LOM metadata from different types of learning objects (e.g. HTML, PDF, PPT, Word). The goal of flexibility was achieved as the system provides a public Web Services API that can be used by third party applications and has a SOA based architecture that makes it possible to extend the framework with new components.

The main components of ALOA are Extractors and Generators. An extractor is responsible for extracting text information from a learning object along with more properties about the learning object. Only one extractor can be defined for each learning object type e.g. html, pdf, ppt, word. A generator is responsible for the actual metatada generation. It uses the output of an extractor and applies data mining techniques to generate one or parts of the metadata. Everyone should be able to easily create a new extrator/generator and plug it into ALOA via the admin interface. A detailed documentation on how to implement a new extractor or generator is available. Please do not hesitate to contact me if you would like to create your own component/service/module or need more information about the ALOA framework.

This thesis was a successful cooperation between Informatik 5, RWTH Aachen University and the computer science department of the Katholieke Universiteit Leuven in the framework of the EU Network of Excellence PROLEARN.


Try it out! Your valuable feedbacks, suggestions, comments, and ideas are welcome!

u-Annotate - An Application for User-Driven Freeform Digital Ink Annotation of E-Learning Content

User-driven annotation of learning content allows the user to interact with it in a flexible and an intuitive manner and hence personalize the content. This interaction, in most cases, follows the “paper and pen” note-taking paradigm. Though not without its advantages, this paradigm does not scale when the courses are delivered online as web pages. These paper notes are not synchronous with the content, hence may seem out of context, they may be misplaced, and don’t make allowances for collaborative learning. In this project we have designed and implemented u-Annotate, a user-driven freeform digital ink annotation application for web e-Learning content which aims at facilitating the learner to annotate the online content with the aid of pen computing devices such as graphic tablets, etc. Learners can freely mark up the content, save the annotations for recall at a later date, as well as share these with other learners.

u-Annotate endeavors to completely replace the paper-and-pen paradigm for the annotating needs. With the aid of alternate input devices, the web page can be thought of as a digital canvas where the user can enter her notes in the marginalia, and draw in a freeform way with a myriad of tools, and enter text via a keyboard into special text boxes.

Flash 8 with ActionScript 2.0 was chosen to be the development tool of choice for the provision of the front-end and rendering/retrieval based on criteria which were a) Support for overlays: SWF files can be transparently overlaid on top of the HTML content. This is the ability to be transparently overlaid in a separate DIV layer with Z Index greater than that of the underlying content. b) Browser Independence: The Flash program once embedded in a web page, can be run on any browser that supports the Macromedia Flash 8 player plug-in. c) Performance: The suitability of the application depends in great part on the basis of how well it performs i.e. speed of rendering, the amount of CPU stress it entails etc. The Flash application performs quite well on these tests.

Project participants are Informatik V, RWTH Aachen University, Germany, bureau42 GmbH in Cologne, Germany and the Open University of Netherlands at Heerlen (NL).

For more informations, please visit the project homepage.