Semantic technologies and learning

The January special issue of Interactive Learning Environments is out right now. Our guest editors have done a great job drawing together 5 excellent papers under the banner of 'Semantic Technologies for Multimedia Enhanced Learning Environments' and for Learning with 'e's readers, here is the editorial in full, with excellent summaries of all the papers by our special issue editors Marco Bertini, Vladan Devedzic, Dragan Gasevic and Carlo Torniai:

Widely available learning material is recognized as a key asset that enables aggregation, provisioning, retrieval, reusability, adaptation, and personalization of educational content. Besides being able to author, publish, discover, and use high-quality learning objects, it is equally important to use multimedia-rich learning objects. Many domains require very advanced content, where different concepts and processes require the use of multimedia (e.g. image, sound, and video) to provide students with a better understanding of concepts under study. This inevitably sets new requirements in multimedia-enhanced learning environments for the advanced representation and creation of learning metadata. The goal is not only to have a richer representation of learning content but it is also important to consider multimedia learning objects in various learning situations where interaction and collaboration are required features. For example, interaction needs to be improved across all the six dimensions of the well-known interactivity triangle with the three main participating nodes of interaction - instructors, students, and content (Anderson & Garrison, 1998). Yet, students are also content creators. This is nicely facilitated by Social Web technologies (e.g. blogs and wikis), which better enable learning environments to support principles of social constructivism. While today user-created multimedia content is a commodity in learning environments, we need to have pedagogical strategies to show how to make the best use of the available technologies. Creative solutions are needed and new perspectives are more than welcome. Just as we can expect learners to easily create and publish multimedia content, we should also facilitate interaction between learners, their peers and educators through multimodal channels of communication and help new users benefit from the experience of previous users of multimedia learning content. Spector (2009) of Google Inc. refers to this phenomenon as “fluidity among the modalities,” where many new modalities will come in addition to the more frequently used ones - text, video, voice, and image.

Scope of the special issue

This special issue analyzes how semantic technologies can be leveraged to address some of the above-mentioned challenges of multimedia-rich learning environments. To perform this analysis, it is first important to define the concept of semantic technologies. Traditionally, the Semantic Web is associated with semantic technologies (Gaevi, Jovanovi, & Devedi, 2007). Ontologies, as the backbone of the Semantic Web defining formally and explicitly represented shared domain conceptualization, are the main way for representing and sharing metadata. Current research in learning technologies has shown that in learning environments Semantic Web technologies can integrate data about learning objects, learning activities and learners captured from various e-Learning systems and tools. Due to the intensive use of Web 2.0 techniques (e.g. collaborative tagging, social networking, mash-ups, and wikis), lightweight representation of semantics and metadata is used in the form of folksonomies, user comments, and ratings. Despite the initial perception that Web 2.0 opposes the Semantic Web, these two efforts are being jointly used to create a common space of semantic technologies (Hendler, 2009). Therefore, semantically enhanced metadata for learning multimedia cannot be considered without the social and interaction context, in which learning constantly happens (Jovanovi, Gaevi, Torniai, Bateman, & Hatala, 2009). Metadata is used to facilitate the discovery and sharing of learning multimedia objects and metadata created through the interaction of learners and educators among themselves and with the learning content.

This special issue solicited papers focused on the use of semantic technologies in multimedia-enhanced learning environments. In this call, we were especially interested in publishing research reports and lessons learned in the following research tasks:

• Ontologies and semantic annotations for multimedia learning objects.


• Collaborative tagging and folksonomies for multimedia learning objects.


• Semantic social networking in multimedia-based learning environments.


• Semantic technologies for enabling pedagogical theories in multimedia-enhanced learning environments.


• Semantic-enhanced learning designs in multimedia-enhanced learning.


• Semantic technologies for personalization and adaptation of multimedia-enhanced learning.


• Semantic-rich service-oriented architectures for multimedia learning environments.


• Semantic multimedia content for (collaborative) mobile learning.

Selected papers

After an enthusiastic response to the open call for papers, followed by a rigorous peer-review process, we are pleased to present five papers addressing some of the indentified research topics. While it would be unrealistic to expect a complete coverage of all research topics due to their vast scope, the paper selection reflects thoroughly the state-of-the-art in this area and some promising research results. More importantly, we can also see many needs for future research, which will hopefully be addressed in the years to come.

In the paper entitled “Automatic generation of tests from domain and multimedia ontologies,” Andreas Papasalouros, Konstantinos Kotis and Konstantinos Kanaris look at the problem of automated generation of quizzes for assessment from domain knowledge. The authors recognized that currently there are many approaches allowing for generating and analyzing tests, but they all rely on text-based content. However, in many different areas, it is important to include multimedia content into the questions asked in quizzes. In their approach, the authors make use of ontologies to represent knowledge of a domain at hand. The domain ontologies are then used together with multimedia annotation ontologies to annotate multimedia learning objects. On top of such annotated multimedia, the authors propose several different strategies for generation of multiple choice questions, where the assessment of students' answers is making use of ontology-based reasoning (i.e. subsumption). Besides applications on text-based content, the authors also demonstrate how their approach can be used for images and argue that the approach can easily be applied to other types of multimedia content. With the use of a prototypical implementation of the proposed approach, the results obtained in the evaluation demonstrate some very promising practical prospects.

Semantic annotation of multimedia learning objects is the topic addressed in the paper entitled “Semantic annotation of video fragments as learning objects: a case study with YouTube videos and the Gene Ontology” by Elena Garca-Barriocanal, Miguel-Angel Sicilia, Salvador Sanchez-Alonso and Miltiadis Lytras. The authors focus their effort on user-generated content (in particular videos posted on YouTube) that can be used as learning material. The need for effective ways to annotate this content is addressed by an annotation tool based on domain ontologies. The generated metadata are then used as a filter for selecting relevant parts of annotated clips as learning objects.

Another paper also focuses on collaborative annotation of multimedia learning content - “A collaborative multimedia annotation tool for enhancing knowledge sharing in CSCL,” by Stephen J.H. Yang, Jia Zhang, Addison Y.S. Su and Jeffrey J.P. Tsai. The authors investigate various annotation techniques (e.g. comments or tags) as instruments helping students develop their critical thinking skills through collaborative learning. In particular, they proposed an architecture based on the use of semantic technologies (for conceptual modelling of collaborative annotations) and web services (for distributed collection and flexible integration of shared annotations). By developing a novel learning environment for collaborative e-Learning and knowledge sharing, using a personalized annotation management system (PAMS 2.0), the authors extensively evaluated the implications of their architecture and approach in a course involving 94 junior university students. The analysis of the collected data indicates that the proposed approach to knowledge sharing helps learners better comprehend their readings and stimulate them ask engaging questions to be discussed with their peers.

The role that semantic technologies can play in reusing and sharing learning resources is well depicted by A. Yessad, C. Faron Zucker, R. Dieng-Kuntz and M.T. Laskri. In their paper entitled “Ontology-based semantic relatedness for detecting the relevance of learning resources,” they describe a novel approach to the computation of the semantic relevance of learning resources to a learning context of a learner. The idea is to compute the relevance between conceptual annotations for the learning resource (built using its role in the learning process and its learning topics) and the concept of interest to the learner. The proposed method offers promising results compared to both semantic measure of similarity and experts ratings.

While it is important to discover some parts of multimedia content, it is also equally important to validate the quality and relevance of the learning content to be used by a learner. In the paper “Constraint modeling for curriculum planning and validation,” Matteo Baldoni, Cristina Baroglio, Ingo Brunkhorst, Nicola Henze, Elisa Marengo and Viviana Patti recognize in authoring of personalized curricula, a gap between learners' traits (e.g. background knowledge or various cognitive traits) and curricula that educational institutions may offer. To address this research challenge, the authors propose a constraint-based technique based on the use of ontologies, model checking principles, and temporal logic to validate whether personal curricula being proposed for each individual learner satisfy the learner's personal traits. The prototypical implementation of the Personal Reader system for education allowed the authors to evaluate their proposed method and to report on some important lessons learned.

Image source by Taylor & Francis