Refine
Year of publication
Document Type
- conference proceeding (article) (177)
- Article (23)
- conference proceeding (presentation, abstract) (7)
- Part of a Book (1)
- conference talk (1)
- conference proceeding (volume) (1)
- Report (1)
- Working Paper (1)
Language
- English (212) (remove)
Is part of the Bibliography
- no (212)
Keywords
- Real-time systems (3)
- Safety (3)
- eye tracking (3)
- higher education (3)
- interface (3)
- reliability analysis (3)
- semantic (3)
- AUTOSAR (2)
- Causality (2)
- Functional Safety (2)
Institute
- Laboratory for Safe and Secure Systems (LAS3) (212) (remove)
Begutachtungsstatus
- peer-reviewed (57)
- begutachtet (11)
In the field of software engineering, graph-based models are used for a variety of applications. Usually, the layout of those graphs is determined at the discretion of the user. This article empirically investigates whether different layouts affect the comprehensibility or popularity of a graph and whether one can predict the perception of certain aspects in the graph using basic graphical laws from psychology (i.e., Gestalt principles). Data on three distinct layouts of one causal graph is collected from 29 subjects using eye tracking and a print questionnaire. The evaluation of the collected data suggests that the layout of a graph does matter and that the Gestalt principles are a valuable tool for assessing partial aspects of a layout.
C is one of the most widely used programming languages - MISRA C is one of the most known sets of coding guidelines for C. This paper examines the usefulness and comprehensibility of the MISRA C:2012 guidelines in an eye tracking study. There, subjects encounter non-compliant code in four different code review settings: with no additional reference, with an actual MISRA C guideline, with a case-specific interpretation of a MISRA C guideline, and with a compliant version of the code. The data collected was analyzed not only in terms of the four presentation styles, but also by dividing the subjects into experience levels based on their semesters of study or years of work experience. Regarding the difference between actual and interpreted guidelines, we found that for interpreted guidelines the error detection rate is higher whereas the duration and frequency of visits to the guideline itself are mainly lower. This suggest that the actual guidelines are less useful and more difficult to understand. The former is contradicted by the subjects’ opinions: when surveyed, they rated the usefulness of the actual guidelines higher.
As humans, we tend to use models to describe reality. Modeling languages provide the formal frameworks for creating such models. Usually, the graphical design of individual model elements is based on subjective decisions; their suitability is determined at most by the prevalence of the modeling language. With other words: there is no objective way to compare different designs of model elements. The present paper addresses this issue: it introduces a systematic approach for evaluating the elements of graph-based modeling languages comprising 14 criteria – derived from standards, usability analyses, or the design theories ‘Physics of Notations’ and ‘Cognitive Dimensions of Notations’. The criteria come with measurement procedures and evaluation schemes based on reasoning, eye tracking, and questioning. The developed approach is demonstrated with a specific use case: three distinct sets of node elements for causal graphs are evaluated in an eye tracking study with 41 subjects.
The dropout rate at universities has been very high for years. Thereby, the inexperience and lack of knowledge of students in dealing with individual learning paths in various courses of study plays a decisive role. Adaptive learning management systems are suitable countermeasures, in which learners’ learning styles are classified using questionnaires or computationally intensive algorithms before a learning path is suggested accordingly. In this paper, a study design for student learning style classification using eye tracking is presented. Furthermore, qualitative and quantitative analyses clarify certain relationships between students’ eye movements and learning styles. With the help of classification based on eye tracking, the filling out of questionnaires or the integration of computationally or cost-intensive algorithms can be made redundant in the future.
This study examines how Klingsieck’s LIST-K questionnaire [22] can be shortened and adapted to the requirements of an online learning management system. In a study with 213 participants, the questionnaire is subjected to an exploitative factor analysis. In a next step, the results are evaluated in terms of their reliability. This process creates a modified factor structure for the LIST-K, comprising a total of eight factors. The reliability of the modified questionnaire is at an α of .770. The shortened version of the LIST-K questionnaire is currently being used on an experimental basis in different courses.
Over the past decade, cars have turned gradually into real cyber physical systems. The collaboration of services between the service-oriented, cross-enterprise vehicle application frameworks has increased to generate novel, smart and complicated vehicle services. Consequently, from an interoperability perspective, semantically mapping of vehicle service component’s interface ontological models emerged as a big research interest in automotive application domain that manipulates several cross-enterprise synergy knowledge applications frameworks. The ontological metamodeling lays the foundation for building semantic bridge and exploring semantic associations between service components’ interface models based on the domain knowledge for semantic interoperability. Also, several semantic quality metrics has been defined over time for the vehicle service interface ontological metamodels. The empirically evaluated values of these metrics can be used to assess progress in cross-enterprise interoperability between the service and the clients’ APIs ontological models in vehicle domain. Despite potential benefits of semantic alignment quality metrics, the effective use of these metrics for vehicle service interface ontologies have proven elusive. Yes, such metrics can be used successfully for quantification, but then they mostly fail to provide adequate annotations in subsequent decision-making in the direction of semantic interoperability and reusability. The effective use of ontology semantic alignment quality metrics is basically hindered in the absence of the meaningful thresholds. In fact, the absence of an effective and meaningful threshold for the semantic similarity measure between various vehicle service interface ontological metamodels, motivates this research work which not only proposes a design approach to an optimized threshold for the semantic similarity metrics but also applies this threshold on few defined semantic alignment quality metrics. This paper also uses a real-world vehicle domain industrial case study to illustrate the design approach.
This study investigates the impact of eye movement modeling examples in Software Engineering education. Software Engineering is a highly visual domain. The daily tasks of a software engineer (e.g., formulating requirements, creating UML diagrams, or conducting a code review) require in many cases the use of certain visual strategies. Although these strategies can be found for experts, it has been observed in different eye tracking studies that students have difficulties in learning and applying them. To familiarize students with these visual strategies and to provide them with a better understanding for the cognitive processes involved, a total of seven eye movement modeling examples was created. The seven eye movement modeling examples cover relevant parts of an introductory Software Engineering lecture; they are focused on typical situations in which visual strategies are applied. The results of a questionnaire-based evaluation shows that students consider the eye movement modeling examples as useful, feel supported in their learning process, and would like to see more use of them in the Software Engineering lecture. Furthermore, the students suggested that eye movement modeling examples should also be used in other lectures.
Nowadays, learning management systems are widely employed in all educational institutions to instruct students as a result of the increasing in online usage. Today’s learning management systems provide learning paths without personalizing them to the characteristics of the learner. Therefore, research these days is concentrated on employing AI-based strategies to personalize the systems. However, there are many different AI algorithms, making it challenging to determine which ones are most suited for taking into account the many different features of learner data and learning contents. This paper conducts a systematic literature review in order to discuss the AI-based methods that are frequently used to identify learner characteristics, organize the learning contents, recommend learning paths, and highlight their advantages and disadvantages.
This paper assesses the relation between personality, demographics, and learning style. Hence, data is collected from 200 participants using 1) the BFI-10 to obtain the participant’s expression of personality traits according to the five-factor model, 2) the ILS to determine the participant’s learning style according to Felder and Silverman, and 3) a demographic questionnaire. From the obtained data, we train and evaluate a Bayesian network. Using Bayesian statistics, we show that age and gender slightly influence personality and that demographics as well as personality have at least a minor effect on learning styles. We also discuss the limitations and future work of the presented approach.
An immense diversity in bottle types requires high accuracy during sorting for recycling purposes by breweries. This extremely complex and time-consuming procedure can result in enormous additional costs for them. This paper presents transfer learning-based algorithms for classifying beer bottle brands using camera images, applicable in individual sorting solutions for different use cases. The problem is tackled using customised EfficientNet, InceptionResNet and VGG models along with an augmented dataset. In addition, a detailed analysis of different model and parameter combinations is performed, enabling tailor-made technologies for specific conditions and resource limitations. In accompanying validations and subsequent tests, a test accuracy of 100% in the recognition of beer brands could be achieved, proving the proposed method fully contributes to the solution of the problem.