Improving Recall and Precision of a Personalized Semantic Search Engine for E-learning

Abstract

The main objective of this paper is to propose and evaluate an architecture that provides, manages, and collects data that permit high levels of adaptability and relevance to the user profiles. In addition, we implement this architecture on a platform called HyperManyMedia. To achieve this objective, an approach for personalized search is implemented that takes advantage of the semantic Web standards (RDF and OWL) to represent the content and the user profiles. The framework consists of the following phases: (1) building the semantic E-learning domain using the known college and course information as concept and sub-concept, (2) generating the semantic user profiles as ontologies, (3) clustering the documents to discover more refined sub-concepts, (4) reranking the user's search results based on his/her profile, and (5) providing the user with semantic recommendations. The implementation of the ontologies models is separate from the design and implementation of the information retrieval system, thus providing a modular framework that is easy to adapt and port to other platforms. Finally, the experimental results show that the user context can be effectively used for improving the precision and recall in E-learning search, particularly by re-ranking the search results based on the user profiles.

Harnessing Background Knowledge for E-Learning Recommendation

“…This often entails using an existing ontology [11,15], or creating a new one [12]. Although ontologies are designed to have a good coverage of their domains, the output is still dependent on the view of its builders, and because of handcrafting, existing ontologies cannot easily be adapted to new domains.…”

“…Recommendation differs from an information retrieval task because with the latter, the user requires some understanding of the domain in order to ask and receive useful results, but in e-learning, learners do not know enough about the domain. Furthermore, the e-learning resources are often unstructured text, and so are not easily indexed for retrieval [11]. This challenge highlights the need to develop suitable representations for learning resources in order to facilitate their retrieval.…”

A Search Engine to Categorize LOs

“…The user's profile must now be updated, in order to trace an evolution of it [8]. [7] also proposes to monitor it periodically, to detect possible shifts of interests.…”

“…[1] proposes to create this profile with a "Bottom-up Pruning" algorithm, that selects the visited LOs from the e-learning tree; [8] suggests to generate the semantics as an ontology [9]. [1] uses than these information to find a recommended cluster for the user.…”

“…[1], [7] and [8] consider fundamental a clusterization of the LOs in macro-groups. Some specific clustering algorithms exist that measure similarities among different objects and group them into categories.…”