...
- As it can be seen from the results, the automatically harvested metadata has poor quality. Not sufficient for a pan-European aggregation portal purposes. It has been proofed and agreed that extensive metadata analysis, common metadata set agreement and engagement with the repository owners (i.e. providers) is needed.
- ARIADNE is only a specification and a set of protocols; a software stack to harvest and index metadata. It has a GoogleSerach-like user interface at the moment.
- The modular architecture helps to add new services to the portal easily.
- Peter (TERENA) said that the initial idea was to compare the MAOR-based and ARIADNE-based implementations, evaluate them with the criteria of a pan-European portal in mind, and implement the TERENA OER pilot portal based on the wining platform. Eli and Giannis commented that the two tools can even be combined as some of the features can nicely complement each other.
- Eli said that e.g., ARIADNE can be used to harvest metadata automatically from the internet (open PMH targets, YouTube channels) while MAOR can be used to harvest metadata from the national repositories engaged with the portal.
- It was agreed to dedicate the next coming meeting of the group to the discussion about how the two platform implementation can be combined.
- Peter (TERENA) will set up the next meeting and send out the open invitation to the TF-Media group.
- After the meeting Michael Stücheli (SWITCH) contacted Peter and offered to harvest metadata from the SWITCHcollection repository. Michael, Eli and Giannis were put in touch.
- In the meantime, Eli and Giannis will try and harvest good quality metadata for both portal implementations from Greece (GRNET), Israel (IUCC), and Switzerland (SWITCH) national repositories.
- Peter (TERENA) will write a news item about the TERENA OER portal pilot and invite other national repositories to participate.
3rd VC preparation
Giannis:
As we discussed during the last meeting I fully support the idea of combining the MAOR and ARIADNE technologies to set up the Terena OER portal. I see the following three options:
- The aggregation of the educational metadata could be performed at a national level following the approach of MAOR while ARIADNE technologies could be used to aggregate all the metadata records at an European/ global level. In this case MAOR instances will be harvested by ARIADNE and the Terena OER portal will be set up on the top of aggregated metadata.
- The ARIADNE based aggregator is used to harvest any available repository (open repositories with video content, youtube etc) and MAOR for national repositories. In this case we need to define how the different metadata schemas will be combined and presented in the Terena OER portal
- ARIADNE infra is used for harvesting, validating and transforming metadata from all national providers and MAOR uses the harvested records for the portal functionalities.
My proposal is to go for the 3rd option and set up an Ariadne labs proposal for this. Of course we can elaborate these ideas during next VM.
Eli:
It is great to see that the Terena-OER initiative is in progress. In my opinion, we can consider a different approach for the Terena OER while supporting the idea of combining MAOR and ARIADNE to set the Terena OER portal. Hence, I suggest the following:
- Reaching a common definition of the metadata fields for the Terena-OER for all NRENs, Universities and content-provider involved, based on the LOM, the Dublin core, or similar standards. This will help us to create a common language for the metadata. This effort can be led by both ARIADNE and MAOR.
- To promote the development of the European educational portal for learning objects repository, Terena-OER portal. This portal will be one step above the national repository (NREN’s or University), and will be able to pull the Metadata to the central Terena-OER by harvesting, web services, OAI repository or batch upload. This will enable us to create the repository that suits the academic user with relevant learning object. The MAOR system can help to promote that stage, with a little modification and development, that can serve as an independent Terena-OER portal.
- After we will have the Terena-OER Portal, with a significant number of our learning objects (recording courses, simulation, animation etc.), we will connect to the GLOBE/ARIADNE/OCW-C or to any global metadata repository. We can connect to those repositories by harvesting, federated search or SQI. It is also worth noting that today, repository such as GLOBE functions as federate of federations. I think that Terena-OER can, and should be a federation.
Steps no. 1 and 2 will help us to promote the project in a way that will be tailored to the Terena end user, NRENs and Universities. In the future, step no. 3 will help us to become a part of the global federation and to increase the number of the learning objects in the repository.
I think that the main difference between the processes and schemes suggested is the expected role of Terena-OER in them. At the beginning, Terena-OER should be the project that combines between the national NERNs in Europe, and in the second stage the project that will link to the international federated searches. Else, Terena-OER will assimilated in a vast repository (such as Ariadne) that includes different educational resources (academia, k-12 and other), without presenting unique features for the Terena end user.
***