Udo Kruschwitz

Nils Constantin Hellwig, Niklas Donhauser, Jakob Fehle et al.

Recent work explored the capabilities of Large Language Models (LLMs) in Aspect-Based Sentiment Analysis (ABSA) through few-shot prompting, requiring substantially fewer annotated examples while achieving notable improvements over zero-shot baselines. However, a performance gap remained compared to models fine-tuned on hundreds of examples, and the computational costs of LLM inference present practical barriers to deployment. We introduce LLM-based Multi-View Prompting (LLM-MvP), which adapts the multi-view principle of considering multiple element orderings to LLM prompting. By combining schema-constrained decoding with a context-free grammar and prefix batching, LLM-MvP achieves performance competitive or superior to fine-tuned approaches while substantially reducing computational overhead. Extensive experiments across five benchmark datasets demonstrate that LLM-MvP closes the gap between few-shot prompting and fine-tuned models, offering a practical and efficient solution for ABSA.

Samy Ateia, Udo Kruschwitz

Commercial large language models (LLMs), like OpenAI's GPT-4 powering ChatGPT and Anthropic's Claude 3 Opus, have dominated natural language processing (NLP) benchmarks across different domains. New competing Open-Source alternatives like Mixtral 8x7B or Llama 3 have emerged and seem to be closing the gap while often offering higher throughput and being less costly to use. Open-Source LLMs can also be self-hosted, which makes them interesting for enterprise and clinical use cases where sensitive data should not be processed by third parties. We participated in the 12th BioASQ challenge, which is a retrieval augmented generation (RAG) setting, and explored the performance of current GPT models Claude 3 Opus, GPT-3.5-turbo and Mixtral 8x7b with in-context learning (zero-shot, few-shot) and QLoRa fine-tuning. We also explored how additional relevant knowledge from Wikipedia added to the context-window of the LLM might improve their performance. Mixtral 8x7b was competitive in the 10-shot setting, both with and without fine-tuning, but failed to produce usable results in the zero-shot setting. QLoRa fine-tuning and Wikipedia context did not lead to measurable performance gains. Our results indicate that the performance gap between commercial and open-source models in RAG setups exists mainly in the zero-shot setting and can be closed by simply collecting few-shot examples for domain-specific use cases. The code needed to rerun these experiments is available through GitHub.

Philipp Hartl, Udo Kruschwitz

The distribution of fake news is not a new but a rapidly growing problem. The shift to news consumption via social media has been one of the drivers for the spread of misleading and deliberately wrong information, as in addition to it of easy use there is rarely any veracity monitoring. Due to the harmful effects of such fake news on society, the detection of these has become increasingly important. We present an approach to the problem that combines the power of transformer-based language models while simultaneously addressing one of their inherent problems. Our framework, CMTR-BERT, combines multiple text representations, with the goal of circumventing sequential limits and related loss of information the underlying transformer architecture typically suffers from. Additionally, it enables the incorporation of contextual information. Extensive experiments on two very different, publicly available datasets demonstrates that our approach is able to set new state-of-the-art performance benchmarks. Apart from the benefit of using automatic text summarization techniques we also find that the incorporation of contextual information contributes to performance gains.

Gregor Donabauer, Udo Kruschwitz

Fake news detection has become a research area that goes way beyond a purely academic interest as it has direct implications on our society as a whole. Recent advances have primarily focused on textbased approaches. However, it has become clear that to be effective one needs to incorporate additional, contextual information such as spreading behaviour of news articles and user interaction patterns on social media. We propose to construct heterogeneous social context graphs around news articles and reformulate the problem as a graph classification task. Exploring the incorporation of different types of information (to get an idea as to what level of social context is most effective) and using different graph neural network architectures indicates that this approach is highly effective with robust results on a common benchmark dataset.

Juntao Yu, Silviu Paun, Maris Camilleri et al.

Although several datasets annotated for anaphoric reference/coreference exist, even the largest such datasets have limitations in terms of size, range of domains, coverage of anaphoric phenomena, and size of documents included. Yet, the approaches proposed to scale up anaphoric annotation haven't so far resulted in datasets overcoming these limitations. In this paper, we introduce a new release of a corpus for anaphoric reference labelled via a game-with-a-purpose. This new release is comparable in size to the largest existing corpora for anaphoric reference due in part to substantial activity by the players, in part thanks to the use of a new resolve-and-aggregate paradigm to 'complete' markable annotations through the combination of an anaphoric resolver and an aggregation method for anaphoric reference. The proposed method could be adopted to greatly speed up annotation time in other projects involving games-with-a-purpose. In addition, the corpus covers genres for which no comparable size datasets exist (Fiction and Wikipedia); it covers singletons and non-referring expressions; and it includes a substantial number of long documents (> 2K in length).

Nils Constantin Hellwig, Jakob Fehle, Udo Kruschwitz et al.

We introduce AnnoABSA, the first web-based annotation tool to support the full spectrum of Aspect-Based Sentiment Analysis (ABSA) tasks. The tool is highly customizable, enabling flexible configuration of sentiment elements and task-specific requirements. Alongside manual annotation, AnnoABSA provides optional Large Language Model (LLM)-based retrieval-augmented generation (RAG) suggestions that offer context-aware assistance in a human-in-the-loop approach, keeping the human annotator in control. To improve prediction quality over time, the system retrieves the ten most similar examples that are already annotated and adds them as few-shot examples in the prompt, ensuring that suggestions become increasingly accurate as the annotation process progresses. Released as open-source software under the MIT License, AnnoABSA is freely accessible and easily extendable for research and practical applications.

Samy Ateia, Udo Kruschwitz

We assessed the performance of commercial Large Language Models (LLMs) GPT-3.5-Turbo and GPT-4 on tasks from the 2023 BioASQ challenge. In Task 11b Phase B, which is focused on answer generation, both models demonstrated competitive abilities with leading systems. Remarkably, they achieved this with simple zero-shot learning, grounded with relevant snippets. Even without relevant snippets, their performance was decent, though not on par with the best systems. Interestingly, the older and cheaper GPT-3.5-Turbo system was able to compete with GPT-4 in the grounded Q&A setting on factoid and list answers. In Task 11b Phase A, focusing on retrieval, query expansion through zero-shot learning improved performance, but the models fell short compared to other systems. The code needed to rerun these experiments is available through GitHub.

Miriam Schirmer, Udo Kruschwitz, Gregor Donabauer

Recent progress in natural language processing has been impressive in many different areas with transformer-based approaches setting new benchmarks for a wide range of applications. This development has also lowered the barriers for people outside the NLP community to tap into the tools and resources applied to a variety of domain-specific applications. The bottleneck however still remains the lack of annotated gold-standard collections as soon as one's research or professional interest falls outside the scope of what is readily available. One such area is genocide-related research (also including the work of experts who have a professional interest in accessing, exploring and searching large-scale document collections on the topic, such as lawyers). We present GTC (Genocide Transcript Corpus), the first annotated corpus of genocide-related court transcripts which serves three purposes: (1) to provide a first reference corpus for the community, (2) to establish benchmark performances (using state-of-the-art transformer-based approaches) for the new classification task of paragraph identification of violence-related witness statements, (3) to explore first steps towards transfer learning within the domain. We consider our contribution to be addressing in particular this year's hot topic on Language Technology for All.

Ignacy Alwasiak, Kene Nnolim, Jaclyn Thi et al.

The DRAGUN Track at TREC 2025 targets the growing need for effective support tools that help users evaluate the trustworthiness of online news. We describe the UR_Trecking system submitted for both Task 1 (critical question generation) and Task 2 (retrieval-augmented trustworthiness reporting). Our approach combines LLM-based question generation with semantic filtering, diversity enforcement using clustering, and several query expansion strategies (including reasoning-based Chain-of-Thought expansion) to retrieve relevant evidence from the MS MARCO V2.1 segmented corpus. Retrieved documents are re-ranked using a monoT5 model and filtered using an LLM relevance judge together with a domain-level trustworthiness dataset. For Task 2, selected evidence is synthesized by an LLM into concise trustworthiness reports with citations. Results from the official evaluation indicate that Chain-of-Thought query expansion and re-ranking substantially improve both relevance and domain trust compared to baseline retrieval, while question-generation performance shows moderate quality with room for improvement. We conclude by outlining key challenges encountered and suggesting directions for enhancing robustness and trustworthiness assessment in future iterations of the system.

Nils Constantin Hellwig, Jakob Fehle, Udo Kruschwitz et al.

We present Self-Consistent Structured Generation (SCSG) for Dimensional Aspect-Based Sentiment Analysis in SemEval-2026 Task 3 (Track A). SCSG enhances prediction reliability by executing a LoRA-adapted large language model multiple times per instance, retaining only tuples that achieve a majority consensus across runs. To mitigate the computational overhead of multiple forward passes, we leverage vLLM's PagedAttention mechanism for efficient key--value cache reuse. Evaluation across 6 languages and 8 language--domain combinations demonstrates that self-consistency with 15 executions yields statistically significant improvements over single-inference prompting, with our system (leveraging Gemma 3) ranking in the top seven across all settings, achieving second place on three out of four English subsets and first place on Tatar-Restaurant for DimASTE.

Nils Constantin Hellwig, Jakob Fehle, Udo Kruschwitz et al.

Training models for Aspect-Based Sentiment Analysis (ABSA) tasks requires manually annotated data, which is expensive and time-consuming to obtain. This paper introduces LA-ABSA, a novel approach that leverages Large Language Model (LLM)-generated annotations to fine-tune lightweight models for complex ABSA tasks. We evaluate our approach on five datasets for Target Aspect Sentiment Detection (TASD) and Aspect Sentiment Quad Prediction (ASQP). Our approach outperformed previously reported augmentation strategies and achieved competitive performance with LLM-prompting in low-resource scenarios, while providing substantial energy efficiency benefits. For example, using 50 annotated examples for in-context learning (ICL) to guide the annotation of unlabeled data, LA-ABSA achieved an F1 score of 49.85 for ASQP on the SemEval Rest16 dataset, closely matching the performance of ICL prompting with Gemma-3-27B (51.10), while requiring significantly lower computational resources.

Niklas Donhauser, Jakob Fehle, Nils Constantin Hellwig et al.

Aspect-Based Sentiment Analysis (ABSA) enables fine-grained opinion analysis by identifying sentiments toward specific aspects or targets within a text. While ABSA has been widely studied for English, research on other languages such as German remains limited, largely due to the lack of high-quality annotated datasets. This paper examines how different annotation sources influence the development of German ABSA. To this end, an existing dataset is re-annotated by experts to establish a ground truth, which serves as a reference for evaluating annotations produced by students, crowdworkers, Large Language Models (LLMs), and experts. Annotation quality is compared using Inter-Annotator Agreement (IAA) and its impact on downstream model performance for different ABSA subtasks. The evaluation focuses on Aspect Category Sentiment Analysis (ACSA) and Target Aspect Sentiment Detection (TASD). We apply State-of-the-Art (SOTA) methods for ABSA, including BERT-, T5-, and LLaMA-based approaches to assess performance differences, spanning fine-tuning and in-context learning with instruction prompts. The findings provide practical insights into trade-offs between annotation reliability and efficiency, offering guidance for dataset construction in under-resourced Natural Language Processing (NLP) scenarios.

Jakob Fehle, Nils Constantin Hellwig, Udo Kruschwitz et al.

Aspect-based Sentiment Analysis (ABSA) extracts fine-grained opinions toward specific aspects within text but remains largely English-focused despite major advances in transformer-based and instruction-tuned models. This work presents a multilingual evaluation of state-of-the-art ABSA approaches across seven languages (English, German, French, Dutch, Russian, Spanish, and Czech) and four subtasks (ACD, ACSA, TASD, ASQP). We systematically compare different transformer architectures under zero-resource, data-only, and full-resource settings, using cross-lingual transfer, code-switching and machine translation. Fine-tuned Large Language Models (LLMs) achieve the highest overall scores, particularly in complex generative tasks, while few-shot counterparts approach this performance in simpler setups, where smaller encoder models also remain competitive. Cross-lingual training on multiple non-target languages yields the strongest transfer for fine-tuned LLMs, while smaller encoder or seq-to-seq models benefit most from code-switching, highlighting architecture-specific strategies for multilingual ABSA. We further contribute two new German datasets, an adapted GERestaurant and the first German ASQP dataset (GERest), to encourage multilingual ABSA research beyond English.

Wan-Hua Her, Udo Kruschwitz

Machine Translation has made impressive progress in recent years offering close to human-level performance on many languages, but studies have primarily focused on high-resource languages with broad online presence and resources. With the help of growing Large Language Models, more and more low-resource languages achieve better results through the presence of other languages. However, studies have shown that not all low-resource languages can benefit from multilingual systems, especially those with insufficient training and evaluation data. In this paper, we revisit state-of-the-art Neural Machine Translation techniques to develop automatic translation systems between German and Bavarian. We investigate conditions of low-resource languages such as data scarcity and parameter sensitivity and focus on refined solutions that combat low-resource difficulties and creative solutions such as harnessing language similarity. Our experiment entails applying Back-translation and Transfer Learning to automatically generate more training data and achieve higher translation performance. We demonstrate noisiness in the data and present our approach to carry out text preprocessing extensively. Evaluation was conducted using combined metrics: BLEU, chrF and TER. Statistical significance results with Bonferroni correction show surprisingly high baseline systems, and that Back-translation leads to significant improvement. Furthermore, we present a qualitative analysis of translation errors and system limitations.

Samy Ateia, Udo Kruschwitz

We present BioRAGent, an interactive web-based retrieval-augmented generation (RAG) system for biomedical question answering. The system uses large language models (LLMs) for query expansion, snippet extraction, and answer generation while maintaining transparency through citation links to the source documents and displaying generated queries for further editing. Building on our successful participation in the BioASQ 2024 challenge, we demonstrate how few-shot learning with LLMs can be effectively applied for a professional search setting. The system supports both direct short paragraph style responses and responses with inline citations. Our demo is available online, and the source code is publicly accessible through GitHub.

Cansu Koyuturk, Emily Theophilou, Sabrina Patania et al.

Large Language Models (LLMs) have transformed human-computer interaction by enabling natural language-based communication with AI-powered chatbots. These models are designed to be intuitive and user-friendly, allowing users to articulate requests with minimal effort. However, despite their accessibility, studies reveal that users often struggle with effective prompting, resulting in inefficient responses. Existing research has highlighted both the limitations of LLMs in interpreting vague or poorly structured prompts and the difficulties users face in crafting precise queries. This study investigates learner-AI interactions through an educational experiment in which participants receive structured guidance on effective prompting. We introduce and compare three types of prompting guidelines: a task-specific framework developed through a structured methodology and two baseline approaches. To assess user behavior and prompting efficacy, we analyze a dataset of 642 interactions from 107 users. Using Von NeuMidas, an extended pragmatic annotation schema for LLM interaction analysis, we categorize common prompting errors and identify recurring behavioral patterns. We then evaluate the impact of different guidelines by examining changes in user behavior, adherence to prompting strategies, and the overall quality of AI-generated responses. Our findings provide a deeper understanding of how users engage with LLMs and the role of structured prompting guidance in enhancing AI-assisted communication. By comparing different instructional frameworks, we offer insights into more effective approaches for improving user competency in AI interactions, with implications for AI literacy, chatbot usability, and the design of more responsive AI systems.

Samy Ateia, Udo Kruschwitz

Agentic Retrieval Augmented Generation (RAG) and 'deep research' systems aim to enable autonomous search processes where Large Language Models (LLMs) iteratively refine outputs. However, applying these systems to domain-specific professional search, such as biomedical research, presents challenges, as automated systems may reduce user involvement and misalign with expert information needs. Professional search tasks often demand high levels of user expertise and transparency. The BioASQ CLEF 2025 challenge, using expert-formulated questions, can serve as a platform to study these issues. We explored the performance of current reasoning and nonreasoning LLMs like Gemini-Flash 2.0, o3-mini, o4-mini and DeepSeek-R1. A key aspect of our methodology was a self-feedback mechanism where LLMs generated, evaluated, and then refined their outputs for query expansion and for multiple answer types (yes/no, factoid, list, ideal). We investigated whether this iterative self-correction improves performance and if reasoning models are more capable of generating useful feedback. Preliminary results indicate varied performance for the self-feedback strategy across models and tasks. This work offers insights into LLM self-correction and informs future work on comparing the effectiveness of LLM-generated feedback with direct human expert input in these search systems.

Gregor Donabauer, Udo Kruschwitz

Pre-training of neural networks has recently revolutionized the field of Natural Language Processing (NLP) and has before demonstrated its effectiveness in computer vision. At the same time, advances around the detection of fake news were mainly driven by the context-based paradigm, where different types of signals (e.g. from social media) form graph-like structures that hold contextual information apart from the news article to classify. We propose to merge these two developments by applying pre-training of Graph Neural Networks (GNNs) in the domain of context-based fake news detection. Our experiments provide an evaluation of different pre-training strategies for graph-based misinformation detection and demonstrate that transfer learning does currently not lead to significant improvements over training a model from scratch in the domain. We argue that a major current issue is the lack of suitable large-scale resources that can be used for pre-training.

Dimitri Ognibene, Gregor Donabauer, Emily Theophilou et al.

The use of large language model (LLM)-powered chatbots, such as ChatGPT, has become popular across various domains, supporting a range of tasks and processes. However, due to the intrinsic complexity of LLMs, effective prompting is more challenging than it may seem. This highlights the need for innovative educational and support strategies that are both widely accessible and seamlessly integrated into task workflows. Yet, LLM prompting is highly task- and domain-dependent, limiting the effectiveness of generic approaches. In this study, we explore whether LLM-based methods can facilitate learning assessments by using ad-hoc guidelines and a minimal number of annotated prompt samples. Our framework transforms these guidelines into features that can be identified within learners' prompts. Using these feature descriptions and annotated examples, we create few-shot learning detectors. We then evaluate different configurations of these detectors, testing three state-of-the-art LLMs and ensembles. We run experiments with cross-validation on a sample of original prompts, as well as tests on prompts collected from task-naive learners. Our results show how LLMs perform on feature detection. Notably, GPT- 4 demonstrates strong performance on most features, while closely related models, such as GPT-3 and GPT-3.5 Turbo (Instruct), show inconsistent behaviors in feature classification. These differences highlight the need for further research into how design choices impact feature selection and prompt detection. Our findings contribute to the fields of generative AI literacy and computer-supported learning assessment, offering valuable insights for both researchers and practitioners.

Nils Constantin Hellwig, Jakob Fehle, Udo Kruschwitz et al.

Aspect sentiment quad prediction (ASQP) facilitates a detailed understanding of opinions expressed in a text by identifying the opinion term, aspect term, aspect category and sentiment polarity for each opinion. However, annotating a full set of training examples to fine-tune models for ASQP is a resource-intensive process. In this study, we explore the capabilities of large language models (LLMs) for zero- and few-shot learning on the ASQP task across five diverse datasets. We report F1 scores almost up to par with those obtained with state-of-the-art fine-tuned models and exceeding previously reported zero- and few-shot performance. In the 20-shot setting on the Rest16 restaurant domain dataset, LLMs achieved an F1 score of 51.54, compared to 60.39 by the best-performing fine-tuned method MVP. Additionally, we report the performance of LLMs in target aspect sentiment detection (TASD), where the F1 scores were close to fine-tuned models, achieving 68.93 on Rest16 in the 30-shot setting, compared to 72.76 with MVP. While human annotators remain essential for achieving optimal performance, LLMs can reduce the need for extensive manual annotation in ASQP tasks.

Aparna Sundar, Tony Russell-Rose, Udo Kruschwitz et al.

As artificial intelligence (AI) becomes increasingly embedded in daily life, designing intuitive, trustworthy, and emotionally resonant AI-human interfaces has emerged as a critical challenge. This editorial introduces a Special Issue that explores the psychology of AI experience design, focusing on how interfaces can foster seamless collaboration between humans and machines. Drawing on insights from diverse fields (healthcare, consumer technology, workplace dynamics, and cultural sector), the papers in this collection highlight the complexities of trust, transparency, and emotional sensitivity in human-AI interaction. Key themes include designing AI systems that align with user perceptions and expectations, overcoming resistance through transparency and trust, and framing AI capabilities to reduce user anxiety. By synthesizing findings from eight diverse studies, this editorial underscores the need for AI interfaces to balance efficiency with empathy, addressing both functional and emotional dimensions of user experience. Ultimately, it calls for actionable frameworks to bridge research and practice, ensuring that AI systems enhance human lives through thoughtful, human-centered design.

Hoai Nam Tran, Udo Kruschwitz

This paper describes our approach (ur-iw-hnt) for the Shared Task of GermEval2021 to identify toxic, engaging, and fact-claiming comments. We submitted three runs using an ensembling strategy by majority (hard) voting with multiple different BERT models of three different types: German-based, Twitter-based, and multilingual models. All ensemble models outperform single models, while BERTweet is the winner of all individual models in every subtask. Twitter-based models perform better than GermanBERT models, and multilingual models perform worse but by a small margin.

Tony Russell-Rose, Philip Gooch, Udo Kruschwitz

Knowledge workers (such as healthcare information professionals, patent agents and recruitment professionals) undertake work tasks where search forms a core part of their duties. In these instances, the search task is often complex and time-consuming and requires specialist expert knowledge to formulate accurate search strategies. Interactive features such as query expansion can play a key role in supporting these tasks. However, generating query suggestions within a professional search context requires that consideration be given to the specialist, structured nature of the search strategies they employ. In this paper, we investigate a variety of query expansion methods applied to a collection of Boolean search strategies used in a variety of real-world professional search tasks. The results demonstrate the utility of context-free distributional language models and the value of using linguistic cues such as ngram order to optimise the balance between precision and recall.

Suzan Verberne, Jiyin He, Gineke Wiggers et al.

Search conducted in a work context is an everyday activity that has been around since long before the Web was invented, yet we still seem to understand little about its general characteristics. With this paper we aim to contribute to a better understanding of this large but rather multi-faceted area of `professional search'. Unlike task-based studies that aim at measuring the effectiveness of search methods, we chose to take a step back by conducting a survey among professional searchers to understand their typical search tasks. By doing so we offer complementary insights into the subject area. We asked our respondents to provide actual search tasks they have worked on, information about how these were conducted and details on how successful they eventually were. We then manually coded the collection of 56 search tasks with task characteristics and relevance criteria, and used the coded dataset for exploration purposes. Despite the relatively small scale of this study, our data provides enough evidence that professional search is indeed very different from Web search in many key respects and that this is a field that offers many avenues for future research.

Thanh Vu, Alistair Willis, Udo Kruschwitz et al.

Recent research has shown the usefulness of using collective user interaction data (e.g., query logs) to recommend query modification suggestions for Intranet search. However, most of the query suggestion approaches for Intranet search follow an "one size fits all" strategy, whereby different users who submit an identical query would get the same query suggestion list. This is problematic, as even with the same query, different users may have different topics of interest, which may change over time in response to the user's interaction with the system. We address the problem by proposing a personalised query suggestion framework for Intranet search. For each search session, we construct two temporal user profiles: a click user profile using the user's clicked documents and a query user profile using the user's submitted queries. We then use the two profiles to re-rank the non-personalised query suggestion list returned by a state-of-the-art query suggestion method for Intranet search. Experimental results on a large-scale query logs collection show that our personalised framework significantly improves the quality of suggested queries.