TAThomas Asselborn

Thomas Asselborn

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Thomas Asselborn has a total of 14 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Treating OCR Output as a Language (TOOL): A Pipeline for OCR Post-Correction

    by Thomas Asselborn, Magnus Bender, Ralf Möller, Sylvia Melzer

    Optical Character Recognition (OCR) enables the digitisation and analysis of large document collections, but recognition errors remain a challenge, particularly for historical documents. This paper presents TOOL (Treating OCR Output as a Language), a modular approach that formulates OCR post-correction as a sequence-to-sequence translation task. A fine-tuned MarianMT encoder–decoder model transforms noisy OCR output into corrected text without modifying the underlying recognition engine. We evaluate TOOL on a collection of historical German vocational education and training documents spanning the period from 1871 to the present day. On 139 held-out test documents, TOOL reduces mean word error rate from 0.792 for Tesseract alone to 0.600, corresponding to a relative reduction of approximately 24.3%. Improvements are also observed in Jaccard bigram similarity, BLEU-4, METEOR, and ROUGE-2, whereas mean normalised Levenshtein distance increases. Comparisons with Hunspell, Phi-4, and Qwen2.5-VL show that the larger language and vision-language models achieve higher accuracy on several metrics. However, TOOL offers a competitive quality–runtime trade-off, with an average end-to-end processing time of 30.12 s per document, compared with 96.57 s for Qwen Transcribed. The results support the feasibility of translation-based OCR post-correction as an additional processing stage in existing OCR workflows, while generalisation to other document collections, OCR engines, and languages remains to be investigated.

  • Open Access

    Ontology-based federated information systems with NLP-enhanced retrieval

    by Sylvia Melzer, Simon Schiff, Franziska Weise, Thomas Asselborn, Meike Klettke, Özgür Lütfü Özçep, Kaja Harter-Uibopuu, Ralf Möller

    Analysing large volumes of humanities data requires considerable time and resources due to the high and varied requirements of researchers in terms of data presentation. In addition, humanities scholars wish to enter their search queries in natural language, quickly synchronising distributed data. Efficient approaches, such as DBoD (DataBasing on Demand), can quickly transfer and insert research data into a database. Natural Language Processing (NLP) is used to make these systems more intuitive and user-friendly, and the use of Ontology-based Data Access (OBDA) ensures that the distributed datasets are mapped semantically correctly. In this article, we explain – using a large dataset from the field of epigraphy – how to build a federated information system that supports natural language queries and how OBDA ensures the semantically correct mapping of distributed datasets. We have significantly reduced the implementation time of information systems and a federated information system from months or years to hours. The research data are project-specific and published in a federated information system, complete with the required styles, allowing direct access to the information. In addition, we have a high usability of the federated information systems with NLP-enhanced retrieval.

  • Open Access

    Generative AI for Interactive Learning in Education

    by Sylvia Melzer, Thomas Asselborn, Ralf Möller

    Integrating Large Language Models (LLMs), such as GPT-4o, into educational settings can enhance teaching, learning, and administrative processes. This article explores the multifaceted applications of LLMs in education, including personalised learning experiences and content creation. It examines the benefits of LLMs, such as improved accessibility, heightened student engagement, and the automation of routine administrative tasks. However, using LLMs are not without challenges, including ethical concerns, technical limitations, and the complexities of integration into traditional educational frameworks. Insights from the University of Hamburg’s lecture: “GenAI in Education, Science, and Society” are highlighted to provide a theoretical and practical perspective on LLM implementations. The paper also discusses future directions for advancing LLMs and fostering interactive collaboration with LLMs to maximise the potential of the interactive learning process in education. This article underscores the importance of thoughtful and responsible deployment and usage of LLMs to improve the educational landscape by addressing opportunities and challenges. We also use the newly defined ‘interactive classroom’ to show how teaching could be organised in the future.


  • Open Access

    Building sustainable information systems and transformer models on demand

    by Thomas Asselborn, Sylvia Melzer, Simon Schiff, Magnus Bender, Florian Andreas Marwitz, Said Aljoumani, Stefan Thiemann, Konrad Hirschler, Ralf Möller

    The growing practice of archiving research data in repositories reflects an upward trend. However, storing data in an RDR (Research Data Repository) does not guarantee that the archived data will always be readily reusable, even if this fulfils the FAIR (Findable, Accessible, Interoperable Reusable) principles. To ensure sustainable RDM (Research Data Management), archiving must consider the future potential for data reuse in a low-threshold fashion. In this article, we demonstrate the utilisation of straightforward methods to implement a so-called warm or hot archiving for research data within an RDR, as opposed to the conventional cold archiving approach. We explore the additional value of using research data in the humanities, emphasising the advantages of maintaining data accessibility and relevance over time. In the humanities, evaluating numerous data sets efficiently is crucial for current and future projects. Reviewing and evaluating relevance is important, particularly when dealing with a substantial number of data sets. Rapid evaluation facilitates profound decisions on the utility of the data for one’s ongoing or upcoming projects. For hot archiving, this means that in addition to the research data, the data should be available in a human-friendly way, i.e., a viewer application to visualise the data should be easily accessible. However, as rapid developments in the IT sector mean that after a few years, it cannot be guaranteed that these viewers or other tools will work, we also show how data can be viewed in a user-specific way via the RDR and how sustainable viewing can be integrated into the RDR. This article presents a generic approach to building sustainable viewers, which we call information systems, or transformer models on demand using data from pre-modern Arabic. In addition, we show that the easy-to-use chatbot ChatGPT can alternatively be context-specifically prepared to deliver more precise results and associated resources in the field of humanities. On the one hand, we have achieved a substantial reduction in the development time of an information system, from months to seconds, as well as the ability to fine-tune BERT (Bidirectional Encoder Representations from Transformers) models without specific knowledge in selecting models or tools. On the other hand, we have developed a chatbot that not only provides project-specific responses but also references the sources.