here you have the message I posted there. It is just a quick overview. The 'good stuff' will be availble soon.
aLFanet addresses adaptation in LMSs following educational standards (IMS-LD, IMS-QTI, IMS-LIP, IMS-CP, IEEE-LOM/IMS-MD) in the full cycle of course creation, usage and maintenance (see http://lttf.ieee.org/learn_tech/issues/october2004/index.html#_Toc88851966). Thus, the goal of aLFanet was not to develop on dotLRN, but we integrated this suite as a service in charge of providing the collaboration part of this iLMS.
The architecture of aLFanet was presented in Madrid at the 'Foro Hispano de dotLRN': Full paper (http://dotlrn.org/file-storage/view/madrid05/05.pdf) and slides (http://dotlrn.org/file-storage/view/madrid05/alfanet.ppt).
Since the key issue at this moment at OpenACS forum is IMS-LD integration in dotLRN, I will just explain how this standard has been integrated and used in aLFanet (and also in dotLRN).
First, regarding the Design Phase, ACE Case has developed an LD Authoring Tool, available at https://sourceforge.net/projects/alfanetat/. This Authoring Tool allows to create IMS-LD compliant courses (levels A, B and C). As a result, an imsmanifiest.xml file is obtained.
At Publication Phase, this imsmanifiest.xml file is imported by an LD Engine and a Collaborative Framework, where the users interact in the Use Phase. The LD enigne is based on Coppercore deveolped at OUNL (http://www.coppercore.org/) and includes a Course Manager and a LD Interpreter. This means that it is in charge of the activity sequencing, roles management, etc. and rest of features defined in design time in IMS-LD. However, IMS-LD does not especify the platform where collaboration services take place. Therefore, to support the collaboration services, we (UNED) chose dotLRN (because of the good experience we had at Innova Group with ACES). So, we have to publish the imsmanifiest.xml in dotLRN as well. What does it mean?
At Design Phase, we have defined two services, FO (Forum) and FS (File storage) which are the services we are taking into account so far. The FO and FS services defined in IMS-LD are created as forums and filestorage objects in dotLRN. Services defined in environments in IMS-LD are created in subgroups in dotLRN.
Nevertheless, to acomplish the research objectives of aDeNu Group (we work on providing runtime/dynamic recommendations based on the users interactions during the course execution) we need a deeper integration between LD and interactions in dotLRN. This integrations should provide us the ontology (which is the IMS-LD structure) that supports our recommendation process. The recommendation process works by learning objectives (we assume that metadata are used to specify the learning objectives for each learning object, activity and service). Thus, we have extended dotLRN model to store these relationships.
I'm afraid that what we have done in dotLRN at aLFanet is not exactly what you are trying to get done in dotLRN now. Undoubtely, it can be of help. And what is more, our research objectives are moving to focus our developments on dotLRN itself. So we would be interested in collaborating in the integration of IMS-LD in dotLRN, contributing with our experience at aLFanet.
Finally, I think it could be of interest for further discussions to copy-paste here some paragraphs of our paper at the Madrid Conference, where the specific contributions to dotLRN at aLFanet are described:
"Thus, code development has been done to integrate dotLRN functionality as interaction services within the Interaction Module. First, the publication mechanism has been implemented to publish IMS-LD courses in the dotLRN environment. Second, extensions to dotLRN packages have been made to integrate them in aLFanet architecture. Third, new services that interact with IMS-LD elements have been developed (e.g. access to the IEEE-LOM metadata of the Learning Objects). Fourth, the outputs of the services have been generated in XML format, so the Presentation Layer can produce adaptation of presentation of dotLRN outputs."
"Moreover, a new package have been implemented in the Interaction Module extending dotLRN development: the imsld-lo package. This package allows the learner to interact with other learners in relation with the learning material of the course and the two components built upon dotLRN, providing comments and ratings to these materials and also accessing to the comments and ratings done by other learners."
"We think that the main technical contributions to the dotLRN community from aLFanet project are the following:
1. Course publication from IMS-LD to dotLRN. The publication process generates new objects such as objectives, learning objects and metadata, environments (as subgroups) and services (forums, file-storage) in existing or extended models in dotLRN.
2. Access to the information related to the learning objects or the course, as well as the integration with dotLRN functionality, so that users can generate comments and ratings associated to them.
3. Service outputs in XML format to perform adaptation of presentation."
We will provide soon access to a demo site where dotLRN community can try and provide feedback on aLFanet.