Linköping Electronic Conference Proceedings
Not a member yet
1113 research outputs found
Sort by
Generative AI and Teachers - For Us or Against Us? A Case Study
Get PDFWe present insightful results of a survey on the adoption of generative artificial intelligence (GenAI) by university teachers in their teaching activities. The transformation of education by GenAI, particularly large language models (LLMs), has been presenting both opportunities and challenges, including cheating by students. We prepared the online survey according to best practices and the questions were created by the authors, who have pedagogy experience. The survey contained 12 questions and a pilot study was first conducted. The survey was then sent to all teachers in multiple departments across different campuses of the university of interest in Sweden: Luleå University of Technology. The survey was available in both Swedish and English. The results show that 35 teachers (more than half) use GenAI out of 67 respondents. Preparation is the teaching activity with the most frequency that GenAI is used for and ChatGPT is the most commonly used GenAI. 59% say it has impacted their teaching, however, 55% say there should be legislation around the use of GenAI, especially as inaccuracies and cheating are the biggest concerns
Green Urban Mobility with Autonomous Electric Ferries: Studies of Simulated Maritime Collisions using Adaptive Stress Testing
Get PDFWith 90% of the world's goods transported by sea vessels, it is crucial to investigate their safety. This is increasingly important as autonomy is being introduced into sea vessels, which transport goods and people. To study the safety of an autonomous ferry's collision avoidance system, we consider the Adaptive Stress Testing (AST) method in this work. AST uses machine learning, specifically reinforcement learning, along with a simulation of a system under test---in our case, an autonomous and electric ferry---and its environment. Whether that simulation is fully or partially observable has implications for the integration into existing engineering workflows. The reason is that the fully observable simulation induces a more complex interface than the partially observable simulation, meaning that the engineers designing and implementing AST need to acquire and comprehend more potentially complex domain knowledge. This paper presents maritime adaptive stress testing (MAST) methods, using the world's first autonomous, electric ferry used to transport people as a case study. Using MAST in multiple scenarios, we demonstrate that AST can be productively utilized in the maritime domain. The demonstration scenarios stress test a maritime collision avoidance system known as Single Path Velocity Planner (SP-VP). Additionally, we consider how MAST can be implemented to test using both fully observable (gray box) and partially observable (black box) simulators. Consequently, we introduce the Gray-Box MAST (G-MAST) and Black-Box MAST (B-MAST) architectures, respectively. In simulation experiments, both architectures successfully identify an almost equal number of failure events. We discuss lessons learned about MAST including the experiences with both the Gray-Box and Black-Box approaches
AI, Data Curation and the Data Readiness of Heritage Collections: Exploring the Swedish Newspaper Archive at KBLab
No full textThe increasing availability of digital material and tools for large-scale computational analysis has produced a growing interest in big data approaches in the humanities and social sciences. However, the vital role of data curation as a precondition for such projects remains underappreciated. This paper details the work of KBLab at the National Library of Sweden in testing AI tools to help curate the digitized newspaper archive and make it more amenable to quantitative, machine learning-based research. It provides a description of the library’s newspaper data to offer orientation to researchers interested in the material, before turning to recount the results of our exploration with automated data curation. It concludes by sketching possible next steps for these exploratory efforts, as well as situating this project within a broader recent turn to conceptualize and prioritize the notion of data readiness. Its principal argument is in drawing attention to data curation as an essential part of any digital research project, not something prior to or external from the research process
Teaching Syntax with CLARIN Corpora and Resources
Get PDFThe recent COVID-19 pandemic has brought online learning to the forefront for learners and teachers. As a consequence, the demand for self-paced and adaptive learning resources has reached unprecedented levels. Fortunately, universities had been using e-learning platforms such as Moodle, or other SCORM-compliant LMS, which has helped make the transition from onsite to on-line learning. However, teachers still have had to design and implement assessment activities in the form of self-correcting activities (true/false, multiple answer questions, mark the words, fill in the blanks questions, etc.). This step has proved to be a major hurdle in the on-site to on-line learning transition, since designing and, most of all, manually editing formative and evaluative assessment activities is a very labour-intensive task. In this article, we present a framework that takes advantage of the corpora and resources available from the LINDAT / CLARIAH-CZ Data & Tools platform in order to generate quizzes and other activities related to syntax. After some background on using NLP for teaching grammar, we present our corpus-to-quiz processing chain, and we outline preliminary results on deploying automatically generated French syntax quizzes in the classroom
Leading by Example: The Use of Generative Artificial Intelligence to Create Pedagogically Suitable Example Sentences
Get PDFSeveral second language acquisition studies have argued in favour of practising vocabulary in authentic contexts. After the tradition of obtaining these usage examples by "invention" (i.e. language experts creating examples based on their intuitions) was superseded by corpus-based approaches (i.e. using dedicated tools to select examples from corpora), the rise of large language models led to a third possible "data source": Generative Artificial Intelligence (GenAI). This paper aims to assess GenAI-based examples in terms of their pedagogical suitability by conducting an experiment in which second language (L2) learners compare GenAI-based examples to corpus-based ones, for L2 Spanish. The study shows that L2 learners find GenAI-based sentences more suitable than corpus-based sentences, with -- on a total of 400 pairwise comparisons -- 265 artificial examples being found most suitable by all learners (compared to 10 corpus-based examples). The prompt type (different zero-shot and few-shot prompts were designed) did not have a noticeable impact on the results. Importantly, the GenAI approach also yielded a number of unsuitable example sentences, leading us to conclude that a "hybrid" method which takes authentic corpus-based examples as its starting point and employs GenAI models to rewrite the examples might combine the best of both worlds
Automatic Text Simplification: A Comparative Study in Italian for Children with Language Disorders
Get PDFText simplification aims to improve the readability of a text while maintaining its original meaning. Despite significant advancements in Automatic Text Simplification, particularly in English, other languages like Italian have received less attention due to limited high-quality data. Moreover, most Automatic Text Simplification systems produce a unique output, overlooking the potential benefits of customizing text to meet specific cognitive and linguistic requirements. These challenges hinder the integration of current Automatic Text Simplification systems into Computer-Assisted Language Learning environments or classrooms. This article presents a multifaceted output that highlights the potential of Automatic Text Simplification for Computer-Assisted Language Learning. First, we curated an enriched corpus of parallel complex-simple sentences in Italian. Second, we fine-tuned a transformer-based encoder-decoder model for sentences simplification. Third, we parameterized grammatical text features to facilitate adaptive simplifications tailored to specific target populations, achieving state-of-the-art results, with a SARI score of 60.12. Lastly, we conducted automatic and manual qualitative and quantitative evaluations to compare the performance of ChatGPT-3.5, and our fine-tuned transformer model. By demonstrating enhanced adaptability and performance through tailored simplifications in Italian, our findings underscore the pivotal role of ATS in Computer-Assisted Language Learning methodologies
Fast Approximation of Shapley Values with Limited Data
Get PDFShapley values have multiple desired and theoretically proven properties for explaining black-box model predictions. However, the exact computation of Shapley values can be computationally very expensive, precluding their use when timely explanations are required. FastSHAP is an approach for fast approximation of Shapley values using a trained neural network (the explainer). A novel approach, called FF-SHAP, is proposed, which incorporates three modifications to FastSHAP: i) the explainer is trained on ground-truth explanations rather than a weighted least squares characterization of the Shapley values, ii) cosine similarity is used as a loss function instead of mean-squared error, and iii) the actual prediction of the underlying model is given as input to the explainer. An empirical investigation is presented showing that FF-SHAP significantly outperforms FastSHAP with respect to fidelity, measured using both cosine similarity and Spearman's rank-order correlation. The investigation further shows that FF-SHAP even outperforms FastSHAP when using substantially smaller amounts of data to train the explainer, and more importantly, FF-SHAP still maintains the performance level of FastSHAP even when trained with as little as 15% of training data
Open Brain AI: An AI Research Platform
Get PDFLanguage assessment is pivotal in identifying therapeutic interventions for speech, language, and communication disorders stemming from neurogenic origins, developmental or acquired, and student performance in the classroom. Traditional assessment techniques, however, are predominantly manual, necessitating extensive time and effort for administration and scoring. Such procedures can exacerbate the stress experienced by patients. In response to these inherent challenges, we introduced Open Brain AI (https://openbrainai.com). This state-of-the-art computational platform leverages advanced AI methodologies, encompassing machine learning, natural language processing, large language models, and automated speech-to-text transcription. These capabilities enable Open Brain AI to autonomously analyze multilingual spoken and written language productions. This work aims to present the development and evolution of Open Brain AI, elucidating its AI-driven language processing components and the intricate linguistic metrics it employs to evaluate the overarching and granular discourse structures. Open Brain AI significantly reduces the workload on researchers, clinicians, and teachers by facilitating rapid and automated language analysis. It allows healthcare and education professionals to optimize their operational processes, reallocating precious time and resources to more personalized user interactions. Moreover, Open Brain AI provides clinicians, researchers, and educators the autonomy to undertake essential data analytics, freeing up more bandwidth to focus on other vital facets of therapeutic intervention and care
Designing digitally-driven integrative interdisciplinarity: Professionalism between protocol and judgement
Get PDFWhile there is a growing discussion of the importance of developing collaborative workflows for interdisciplinary research within DH, there is a lack of blueprints and consideration of specific expertise. This paper conceptualizes the practice of what we tentatively call digitally-driven integrative interdisciplinary project design in order to highlight a certain professional practice for integrating collaboration between technical expertise and traditional HSS researchers when developing research project applications, digital resources, etc. We begin by highlighting the need for protocol for workflow- oriented approaches to integrative interdisciplinary collaboration, but also an embodied expertise in need of being put into focus in discussions of integrative workflows within digital humanities. Then, we argue that judgement is also a crucial but often overlooked part of the professionalism involved. We conclude by discussing how to further develop the conceptualization of interdisciplinary digital project design and the expertise involved
Accessing centuries of documentation - Resources to improve access to Swedish rock art documentation and metadata
Get PDFThe archive of rock art documentation maintained by SHFA provides a valuable resource to archaeologists and others who study rock art. The archive includes images of rock art documentation, sites, and the documentation process, from the 17th century to the more recent high resolution 3D recording and visualizations. In the last few years, GRIDH, in collaboration with SHFA, have begun to improve access to the archive through a Django-based solution and new digital resources. In this paper, we introduce the images in the archive, provide details on the new digital resources, and reflect on how the new website will impact data availability and rock art research