DOI: 10.3897/jucs.137103

Authors: Hendrik Hendrik, Silmi Fauziati, Adhistya Erna Permanasari

Abstract: Knowledge graphs are a powerful way to represent and organize complex knowledge. They are used in many fields, like healthcare and finance. They allow for more insightful decision-making and discoveries. However, the quality of knowledge graphs depends heavily on their sources. Current methods for evaluating these sources are often slow and not scalable. They struggle to keep up with the large amount of online information. We created a new tool to address this problem. Our tool uses Large Language Models (LLMs) to assess online sources quickly. It evaluates websites based on credibility, relevance, content quality, coverage, comprehensiveness, and accessibility. We tested our tool on Halal tourism websites in Japan. We compared LLM evaluations with human expert judgments. Our comprehensive analysis revealed that certain LLM models, particularly GPT-3.5-turbo, GPT-4, and Mixtral-8x7B-Instruct-v0.1, showed strong correlation with human evaluations. Using a temperature setting of 0.4, these models demonstrated consistent and reliable performance across multiple evaluation runs. Our structured evaluation framework, incorporating weighted criteria validated through both expert input and statistical analysis, provides a robust foundation for automated source assessment. While some models showed varying performance across different criteria, our findings suggest that careful model selection and potential ensemble approaches could optimize evaluation accuracy. Our work contributes significantly to improving knowledge graph construction by demonstrating the viability of LLM-based source evaluation, while also identifying key areas for future research in scalability, cross-domain validation, and automated optimization.

HTML

PDF

DOI: 10.3897/jucs.2020.049

Authors: Gareth Young, Sam Stehle, Burcin Walsh, Egess Tiri

Abstract: irtual reality (VR), as an informative medium, possesses the potential to engage students with immersive, interactive, and informative experiences. When presented in VR, immersive virtual environments (IVEs) can provide three-dimensional visual simulations that can be used to inform students about concepts in specific contexts that would be near impossible to achieve with more traditional teaching methodologies. It is proposed that existing learning frameworks can benefit from exploring the modalities of interaction that are presently afforded via VR from the experiential perspectives of the students. An evaluation is presented that focused on the appraisal of student experiences of immersive technologies as applied in a higher education context, specifically in the use of VR for the exploration of geomorphology theory by physical geography students. This research supports further development of the immersive learning discipline from three different perspectives. First, an empathy mapping method was applied to visualize student experiences and externalize our observed knowledge of student users for creating a shared understanding of their needs and to aid in lesson planning decision making when using VR in the classroom. Second, student experiences were captured using a technology-focused user experience questionnaire to obtain student attitudes immediately post-task. Finally, to assist teachers with the creation of a student-centered lesson plans that incorporate VR in the classroom, eight heuristic guidelines (focus, provocation, stimulation, collaboration, control, digital life, learner skills, multimodal experience) were developed. It is proposed that these findings can be used to provide support for the use of mixed reality and immersive virtual environments in learning that encompass the challenges faced by students and the interdisciplinary education community at large.

HTML

XML

PDF

DOI: 10.3217/jucs-025-12-1608

Authors: Peter De Lange, Petru Nicolaescu, Jan Benscheid, Ralf Klamma

Abstract: In both formal and informal learning scenarios, storytelling is a mean towards acquiring and sharing knowledge. On theWeb, user generated content and digital representations of real-world artifacts contribute to the story's expressiveness. 3D objects are a good example of such representations, as they capture and distribute rich and complex information. Currently, a bridge between digital storytelling, social Web 2.0 features - such as tagging - and 3D objects is missing. We present an approach for the collaborative creation of non-linear stories in near real-time, centered around 3D objects. We use a metamodel, which acts as a basis for collaborative story creation. Stories are directly linked to 3D objects being attached to camera views and perspectives on a certain object. We instantiate story viewers from the collaboratively authored stories and use text and multimedia to annotate and browse the 3D objects. Our conducted evaluation proves the feasibility of the approach and promises good results in applying collaborative storytelling for 3D object browsing in order to scaffold learning.

HTML

XML

PDF

DOI: 10.3217/jucs-025-05-0541

Authors: Guillermo Santamaría-Bonfil, Guillermo Escobedo-Briones, Miguel Perez-Ramírez, Gustavo Arroyo-Figueroa

Abstract: Linemen training is mandatory, complex, and hazardous. Electronic technologies, such as virtual reality or learning management systems, have been used to improve such training, however these lack of interoperability, scalability, and do not exploit trace data generated by users in these systems. In this paper we present our ongoing work on developing a Learning Ecosystem for Training Linemen in Maintenance Maneuvers using the Experience API standard, Big Data components, and Learning Analytics. The paper describes the architecture of the ecosystem, elaborates on collecting learning experiences and emotional states, and applies analytics for the exploitation of both, legacy and new data. In the former, we exploit legacy e-Learning data for building a Domain model using Text Mining and unsupervised clustering algorithms. In the latter we explore self-reports capabilities for gathering educational support content, and assessing students emotional states. Results show that, a suitable domain model for personalizing maneuvers linemen training path can be built from legacy text data straightforwardly. Regarding self reports, promising results were obtained for tracking emotional states and collecting educational support material, nevertheless, more work around linemen training is required.

HTML

XML

PDF

DOI: 10.3217/jucs-024-01-0043

Authors: Erick Nilsen Pereira De Souza, Daniela Claro, Rafael Glauber

Abstract: Open information extraction (Open IE) discovers facts as triples of relationships in texts. A major challenge to Open IE task is to reduce the proportion of invalid extractions. Current methods based on a set of specific features eliminate many inconsistent and incomplete facts. However, these solutions have the disadvantage of being highly language-dependent. This dependence arises from the difficulty in finding the most representative set of features, considering the peculiarities of each language. These solutions require extensive training sets, usually produced with the aid of a specialized linguistic knowledge. Furthermore, although linguistic knowledge resources are common in English, they are scarce in most other languages. Therefore, we propose a method for classifying extracted facts based on the similarity of grammatical structures, which builds models from morphological structures contained in the extraction through the application of algorithms for the detection of isomorphism in sub-graphs. In particular, Portuguese was chosen for the implementation and validation of the proposed approach as it is one of the languages that lack this type of resource.

HTML

XML

PDF

DOI: 10.3217/jucs-022-03-0360

Authors: Gui-Xun Luo, Yun Liu, Zhi-Yuan Zhang

Abstract: In this paper, based on abundant data from Sina Weibo, we perform a comprehensive and in-depth empirical analysis of repostings and draw some conclusions. First, in regards to quantity, reposting takes up a large proportion of daily microblog activity. Second, the depth of repostings follows an exponential distribution and the first three orders of repostings hold 99 percent of the total amount of reposting, which provides an important foundation for solving the question of Influence Maximization. Third, the time interval for repostings also obeys exponential distribution. Therefore, we have built a dynamic information propagation model in terms of conclusions drawn from Weibo data and the Continuous-Time Markov Process. Due to the basis of the temporal network, our proposed model can change with the time and structure of a network, thus giving it good adaptability and predictability as compared to the traditional information diffusion model. From the final simulation results, our proposed model achieves a good predictive effect.

HTML

XML

PDF

DOI: 10.3217/jucs-021-08-1086

Authors: Lluís Safont, Sergi Villagrasa, David Fonseca, Ernest Redondo

Abstract: Using enhanced learning technologies (TEL) including immersive virtual reality environments, we are seeking to achieve a new way of assessing subjects of 3D arts. We have developed a project based on Scenario Centered Curriculum (SCC), where the students have to think, design, convey, validate, and build a civil project using new technologies that help in the assessment process. We have used gamification techniques and game engines to evaluate planned tasks in which students can demonstrate the skills they developed in the scenarios. The assessment is integrated in the creation of a 3D complex model focused on the construction of a building in a virtual space. This whole process will be carried out by gamification techniques to embed the assessment of the 3D models with the objective of improving students learning.

HTML

XML

PDF

DOI: 10.3217/jucs-018-08-1048

Authors: Victor Ströele, Geraldo Zimbrão, Jano Souza

Abstract: Social networks are dynamic social structures consisting of individuals or organizations, usually represented by nodes tied by one or more relationship type. Analyzing these structures enables us to detect several inter and intra connections between people in and outside their organizations. In this context, we construct a multi-relational scientific social network where researchers may have four different types of relationships with each other. We adopt some criteria such as relationship age in order to assign a weight to relationships and to enable the modeling of a scientific social network as close as possible to reality. Using clustering techniques with maximum flow measure, we identify the social structure and research communities in a way that allows us to evaluate the knowledge flow in the Brazilian scientific community.

HTML

XML

PDF

DOI: 10.3217/jucs-018-03-0327

Authors: Mike Tissenbaum, Michelle Lui, James Slotta

Abstract: This paper introduces a series of iterative designs that investigate how the aggregation and visualization of student-contributed work can support collaborative problem solving in the domain of physics. We investigate how new technologies can enable students to contribute to a shared knowledge base, working across contexts: in class, at home, and in a specialized "smart classroom" environment. We explore how student data can be provided to the teacher before class, in support of planning the next day's lesson, and during class, to help the teacher orchestrate class activities and respond to student needs. Our work builds upon the research tradition of knowledge communities and inquiry learning to inform its design of materials and activities that support productive collaborative interactions for learners. We are also guided by the recent literature on scripting and orchestration to define curricular activities that bridge home and school environments, leveraging a digital platform that includes Web 2.0 features to guide structured collaborations. This paper reports on a design-based research program in which the development of the curriculum and technology platform is informed by successive cycles of design, enactment, analysis, and re-design. The paper will review our efforts through three successive design cycles, exploring the evolution of our own "smart classroom curriculum" for high school physics. For each iteration, we present our design goals, the resulting curriculum and technology, the student learning outcomes, and our evaluation that informs the next iteration. We end with a description of our current design, and discuss the goals and directions of our future efforts.

HTML

XML

PDF

DOI: 10.3217/jucs-017-12-1743

Authors: George Sielis, Christos Mettouris, George Papadopoulos, Aimilia Tzanavari, Roger Dols, Quintin Siebers

Abstract: The development of methods that can enhance the creativity process is becoming a continuous necessity. Through the years several researchers modelled and defined creativity focusing to the psychological aspect of the topic. More recent researchers approach creativity as a computerized process by simulating it within creativity support tools (CST). This article supports that usage of context aware recommender system, in creativity support tools and more specifically, collaborative creativity support tools (CCST) can enhance creativity process. In this work we focus on the development of a context awareness recommender system and look into how such a system can be useful for the creativity process, through preliminary evaluation results in regards to its usefulness and usability.

HTML

XML

PDF

DOI: 10.3217/jucs-017-12-1690

Authors: Peter Dolog, Frederico Durao, Karsten Jahn, Yujian Lin, Dennis Peitersen

Abstract: In this paper we describe a concept of the recommender system for collaborative real time web based editing in the context of creativity sessions. The collaborative real time editing provides creativity teams of which members are physically distributed with an emulation of the synchronous collaboration where presence of the team members is required simultaneously (e.g., brainstorming, meetings). The concept of recommendation is based on matchmaking the currently performed activities at the user interface and external linked open data provided through SPARQL endpoints. The real time propagation of the changes in editor and recommendation is achieved by reverse AJAX and observer pattern. An experiment in the area of the creativity domain shows that the recommendation in collaborative real time editing activities are useful in task performance, guidance, and inspiration.

HTML

XML

PDF

DOI: 10.3217/jucs-016-20-3102

Authors: Bernhard Thalheim

Abstract: Conceptual modelling is a widely applied practice and has led to a large body of knowledge on constructs that might be used for modelling and on methods that might be useful for modelling. It is commonly accepted that database application development is based on conceptual modelling. It is however surprising that only very few publications have been published on a theory of conceptual modelling. Modelling is typically supported by languages that are well-founded and easy to apply for the description of the application domain, the requirements and the system solution. It is thus based on a theory of modelling constructs. At the same time, modelling incorporates a description of the application domain and a prescription of requirements for supporting systems. It is thus based on methods of application domain gathering. Modelling is also an engineering activity with engineering steps and engineering results. It is thus engineering. The first facet of modelling has led to a huge body of knowledge. The second facet is considered from time to time in the scientific literature. The third facet is underexposed in the scientific literature. This paper aims in developing principles of conceptual modelling. They cover modelling constructs as well as modelling activities as well as modelling properties. We first clarify the notion of conceptual modelling. Principles of modelling may be applied and accepted or not by the modeler. Based on these principles we can derive a theory of conceptual modelling that combines foundations of modelling constructs, application capture and engineering. A general theory of conceptual modelling is far too comprehensive and far too complex. It is not yet visible how such a theory can be developed. This paper therefore aims in introducing a framework and an approach to a general theory of conceptual modelling. We are however in urgent need of such a theory. We are sure that this theory can be developed and use this paper for the introduction of the main ingredients of this theory.

HTML

XML

PDF

DOI: 10.3217/jucs-016-19-2841

Authors: Panayiotis Kyriakou, Ioannis Hatzilygeroudis, John Garofalakis

Abstract: Intelligent Tutoring Systems (ITSs) constitute a popular type of intelligent educational systems. Domain Knowledge (DK) is a basic part of an ITS and usually includes information about the concepts the ITS is dealing with and the teaching material itself. The teaching material consists of a set of learning objects (LOs). A LO is described by a data set called its metadata. Concepts are usually organized in a network, called a concept network (or map). Each concept is associated with a number of LOs. Existing tools for managing DK mainly deal with either LOs or concepts, but not with connecting them. In this paper, we present a tool for managing both types of information in DK: creating and editing a concept network and LO metadata as well as connecting them. Additionally, the tool can produce corresponding XML descriptions for each LO metadata. Finally, it provides facilities for helping tutors in organizing and composing their lessons. A small scale evaluation has shown more than satisfactory acceptability of the tool.

HTML

XML

PDF

DOI: 10.3217/jucs-015-13-2701

Authors: Thanh-Quang Chu, Alexis Drogoul, Alain Boucher, Jean-Daniel Zucker

Abstract: Efficient response to natural disasters has an increasingly important role in limiting the toll on human life and property. The work we have undertaken seeks to improve existing models by building a Decision Support System (DSS) of resource allocation and planning for natural disaster emergencies in urban areas. A multi-agent environment is used to simulate disaster response activities, taking into account geospatial, temporal and rescue organizational information. The problem we address is the acquisition of situated expert knowledge that is used to organize rescue missions. We propose an approach based on participatory design and interactive learning which incrementally elicits experts’ preferences by online analysis of their interventions with rescue simulations. An additive utility functions are used, assuming mutual preferential independence between decision criteria, as a preference for the elicitation process. The learning algorithm proposed refines the coefficients of the utility function by resolving incremental linear programming. For testing our algorithm, we run rescue scenarios of ambulances saving victims. This experiment makes use of geographical data for the Ba-Dinh district of Hanoi and damage parameters from well-regarded local statistical and geographical resources. The preliminary results show that our approach is initially confident in solving this problem.

HTML

XML

PDF

DOI: 10.3217/jucs-014-17-2758

Authors: Brent Martin, Antonija Mitrovic, Pramuditha Suraweera

Abstract: Authoring ITS domain models is a difficult task requiring many skills. We explored whether modeling ontology reduces the problem by giving the students of an e-learning summer school the task of developing the model for a simple domain in under sixty minutes using ontology. Some students also used our tool to develop a complete tutor in around eight hours, which is much faster than they could be expected to author the system without the tool. The results suggest this style of authoring can lead to very rapid ITS development. We further extend the ontological approach with domain schema: high-level abstractions that describe the semantics of the domain model for a class of domains. Using domain schema reduces the authoring effort to one of describing only those aspects that are unique to this particular domain, and enables the ontology-based approach to model domains with different semantic requirements.

HTML

XML

PDF