The intention of this deliverable is to describe the implications from the last year’s
evaluation study of ASPL-v1 and how we addressed the identified issues. We briefly
summarize the ASPL functionality and the outcomes of the evaluation, and proceed
with analyzing the learning task we set for the ASPL in the previous deliverables. In
particular, we use the analysis to highlight how the identified gaps and shortcomings
could be addressed so that the benefits and added value of the Semantic Web
approach in designing the ASPL services becomes clearer and more explicit.

In particular we discuss and clarify the position with respect to the amendments made
to the services supporting query refinement/expansion on the retrieved data (see
section about “Improved sensemaking support…”) Another amendment where we
believe benefits will be more visible is described in section “Exploratory navigation
through…”, where we illustrate how automatically harvested correlations relations
could be used to navigate the space of query results.

Additionally, we present our new service about question-based learning that was
implemented since the evaluation of the old version of ASPL. This service is now
fully integrated in the framework. In a similar role, we discuss other recommendation
and data processing services, which are candidates for the inclusion into ASPL in the
next period.

To complement the implementation work on ASPL services, the document briefly
addresses and discusses a problem-centric approach to constructing ontologies for
learning and educational support, and finally, we also highlight some initial steps
towards formulating a more principled method for evaluating the current version of
ASPL for its effectiveness (rather than its rate of recall).