05 Fakultät Informatik, Elektrotechnik und Informationstechnik
Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/6
Browse
201 results
Search Results
Item Open Access Interactive exploration and model analysis for coreference annotation(2013) Gärtner, MarkusI present the design and implementation of an interactive visualization- and exploration-framework for coreference annotations. It is designed to meet the needs of multiple different users on a modern and multifaceted graphical exploration tool. To demonstrate its suitability for these various needs I outline several use cases and how the framework can help users in their individual tasks. It offers the user different views on the data with additional functionality to compare several annotations. Complex analysis of annotated corpora is supported by means of a search engine which lets the user construct queries both in a graphical and textual form. Both qualitative and quantitative result breakdowns are available and the implementation features specialized visualizations to aggregate complex search results. The framework is extensible in many ways and can be customized to handle additional data formats.Item Open Access Challenges of computational social science analysis with NLP methods(2022) Dayanik, Erenay; Padó, Sebastian (Prof. Dr.)Computational Social Science (CSS) is an emerging research area at the intersection of social science and computer science, where problems of societal relevance can be addressed by novel computational methods. With the recent advances in machine learning and natural language processing as well as the availability of textual data, CSS has opened up to new possibilities, but also methodological challenges. In this thesis, we present a line of work on developing methods and addressing challenges in terms of data annotation and modeling for computational political science and social media analysis, two highly popular and active research areas within CSS. In the first part of the thesis, we focus on a use case from computational political science, namely Discourse Network Analysis (DNA), a framework that aims at analyzing the structures behind complex societal discussions. We investigate how this style of analysis, which is traditionally performed manually, can be automated. We start by providing a requirement analysis outlining a roadmap to decompose the complex DNA task into several conceptually simpler sub-tasks. Then, we introduce NLP models with various configurations to automate two of the sub-tasks given by the requirement analysis, namely claim detection and classification, based on different neural network architectures ranging from unidirectional LSTMs to Transformer based architectures. In the second part of the thesis, we shift our focus to fairness, a central concern in CSS. Our goal in this part of the thesis is to analyze and improve the performances of NLP models used in CSS in terms of fairness and robustness while maintaining their overall performance. With that in mind, we first analyze the above-mentioned claim detection and classification models and propose techniques to improve model fairness and overall performance. After that, we broaden our focus to social media analysis, another highly active subdomain of CSS. Here, we study text classification of the correlated attributes, which pose an important but often overlooked challenge to model fairness. Our last contribution is to discuss the limitations of the current statistical methods applied for bias identification; to propose a multivariate regression based approach; and to show that, through experiments conducted on social media data, it can be used as a complementary method for bias identification and analysis tasks. Overall, our work takes a step towards increasing the understanding of challenges of computational social science. We hope that both political scientists and NLP scholars can make use of the insights from this thesis in their research.Item Open Access Supervised semantic proximity noise and disagreement detection(2024) Choppa, TejaswiThe quality and reliability of annotated data are crucial for the development of Machine Learning models. In this work, we particularly focus on word sense annotation in context (a.k.a. Word-in-Context, WiC). WiC datasets in real-world contexts often exhibit significant disagreement. As a result, information is lost when instances are discarded during the creation of the gold label by adjudicating the annotations through majority or median judgment. Recent advancements have sought to address this issue by incorporating disagreement data through novel label aggregation methods (Uma et al., 2022). Modeling this disagreement is important because, in a real-world scenario, we often do not have clean data. We need to predict on samples where high disagreement is expected and which are inherently difficult to categorize. Predicting disagreement can help detect or filter highly complex samples. Through this thesis, we aim to build machine learning models that predict human disagreement in annotated text instances. Moreover, we focus on data with noise instances where annotators cannot confidently assign a label or the data does not fit predefined categories. We aim to measure both disagreement and noise, as they both stem from a common source: ambiguity. By modeling these aspects, we aim to design modeling approaches that predict not only the semantic proximity label but also the annotator disagreement, as well as data noisiness.Item Open Access KGGLDM : Knowledge Graph Guided Diffusion Models for advanced learning(2024) Gupta, AkshatThis thesis explores a novel approach by bridging the gap of diffusion modeling and knowledge graphs, unveiling a potentially groundbreaking direction that serves as the central theme of this work. We propose incorporating knowledge graph guidance into LDM models to augment precise control over sample generation using domain conceptual knowledge.Item Open Access Understanding and adapting to user mental models in task-oriented dialog(2026) Vanderlyn, Lindsey; Vu, Ngoc Thang (Prof. Dr.)Item Open Access Exploring the effects of enriched English language input on language model efficiency(2024) Zeller, TomRecent years have seen the advent of large-scale language modeling as exemplified by transformer-based models like GPT or variants of the BERT architecture. These models, which are trained on massive datasets and using compute unattainable by actors that are not of the scale of the biggest tech companies, have shown impressive feats of syntactic and semantic understanding. Naturally, interest has risen in making these models more efficient, in terms of compute as well as data requirements. Research in this area can be seen as primarily motivated by two factors: reducing the barrier for smaller actors like research institutes or end consumers to train and execute state-of-the-art models, as well as reducing the carbon footprint of these models. To achieve this goal, model compression techniques like quantization, pruning or distillation are utilized. This work aims to explore a different, less model-centric and more data-centric approach: Modifying the training and inference data, by enriching it with syntactic and semantic information. To this end, a lexical resource is created which maps English words to a form where individual characters represent values of a range of semantic and syntactic features, providing lexical information that is accessible to all model types that operate on tokens at the sub-word or character-level. Different features and methods of representation are discussed, and their effect on model performance is evaluated by pretraining a small GPT-family model and fine-tuning on downstream tasks of the SuperGLUE benchmark. Given a fixed amount of data and compute, the experiments show a performance advantage for a character-level model trained using the enriched data.Item Open Access Comparison of distributional and visual nearest neighbors(2025) Naber, SvenThis thesis investigates how semantic concepts are represented across textual and visual embedding spaces, focusing on the abstract-concrete continuum. Using 5,448 English nouns and their embeddings from both distributional language models (e.g., Word2Vec, GloVe) and vision models (e.g., ViT, DINOv2, CLIP), it compares neighborhood structure via a normalized alignment score (NAS). Results show that alignment is primarily driven by input modality rather than model architecture, with strong local overlap for concrete concepts and more diffuse agreement for abstract ones. Mean aggregation of image embeddings improves visual consistency but cannot fully bridge modality-specific limitations. The findings provide a starting point for further exploration of semantic spaces.Item Open Access More reliable retrieval augmented generation for domain-specific question-answering through domain-infused soft prompts(2025) Nassar, ZeinaTransformer-based large language models (LLMs) have revolutionized the field of artificial intelligence, enabling advancements in various applications due to their exceptional reasoning capabilities. However, these models often face challenges such as hallucination, outdated knowledge, and non-transparent reasoning, which limit their reliability in critical tasks like question answering (QA). QA systems, especially in domain-specific contexts like car manuals, require high accuracy and reliability to ensure user safety. Achieving this is complicated by the need for precise retrieval and interpretation of domain-specific information, which can be hindered by unfamiliar keywords or ambiguous questions. Open-domain QA relies on external knowledge repositories and typically follows a retriever-reader framework to locate and process evidence. In contrast, domain-specific QA often lacks sufficient gold-standard datasets, making it essential to explore techniques like retrieval-augmented generation (RAG), which combines retrieved context with the question as input to LLMs. While RAG improves grounding, hallucinations still occur when models rely on pre-trained knowledge over retrieved evidence. This thesis investigates the role of prompting techniques in improving faithfulness and reducing hallucinations in domain-specific QA. By testing domain-specific and domain-agnostic discrete prompts as well as soft prompting methods, this work aims to identify strategies for generating more accurate and grounded responses. The study addresses key research questions on the effectiveness of domain-specific information in prompts, the best way to incorporate such information, and whether dynamic soft-prompts outperform static ones in domain-specific QA scenarios. The findings aim to contribute to building more reliable and factual QA systems.Item Open Access Predicting sentiment about places of living(2017) Liu, FeifeiNowadays studies about the quality of life in major cities are often published in the daily news. These contain ranked list according to the quality of living with indicators representing various aspects. Typical indicators are crime level, transport, health care etc. Along with the flourishing of different social medias, a huge amount of information could be collected from the Internet. Moreover, machine learning as a branch of artificial intelligence becomes more and more prominent. The recent advances in machine learning had found usage in a wide range of applications. One of such application is that of text categorization and sentiment analysis. Relying on these conditions, this thesis aims to create a classifier to predict the sentiment about places of living. In this thesis a ranking list of cities of Mercer is taken use. As a result of the quality of living survey 230 cities of the world are ranked in the list. Text form information of microblogging is chosen as our testbed. Specifically, tweets, microblogging messages from the popular website Twitter, are studied. The tweets chosen for this study are those about cities living standard and contain rich sentiment information. Classification label is assigned to cities under study by their position in the ranking list. After sentiment related features are extracted, machine learning techniques are then applied on the collected tweets. As a result, a classifier with a strong baseline for predicting sentiment about places of living is trained using logistic regression model.Item Open Access Advances in clinical voice quality analysis with VOXplot(2023) Barsties von Latoszek, Ben; Mayer, Jörg; Watts, Christopher R.; Lehnert, BernhardBackground: The assessment of voice quality can be evaluated perceptually with standard clinical practice, also including acoustic evaluation of digital voice recordings to validate and further interpret perceptual judgments. The goal of the present study was to determine the strongest acoustic voice quality parameters for perceived hoarseness and breathiness when analyzing the sustained vowel [a:] using a new clinical acoustic tool, the VOXplot software. Methods: A total of 218 voice samples of individuals with and without voice disorders were applied to perceptual and acoustic analyses. Overall, 13 single acoustic parameters were included to determine validity aspects in relation to perceptions of hoarseness and breathiness. Results: Four single acoustic measures could be clearly associated with perceptions of hoarseness or breathiness. For hoarseness, the harmonics-to-noise ratio (HNR) and pitch perturbation quotient with a smoothing factor of five periods (PPQ5), and, for breathiness, the smoothed cepstral peak prominence (CPPS) and the glottal-to-noise excitation ratio (GNE) were shown to be highly valid, with a significant difference being demonstrated for each of the other perceptual voice quality aspects. Conclusions: Two acoustic measures, the HNR and the PPQ5, were both strongly associated with perceptions of hoarseness and were able to discriminate hoarseness from breathiness with good confidence. Two other acoustic measures, the CPPS and the GNE, were both strongly associated with perceptions of breathiness and were able to discriminate breathiness from hoarseness with good confidence.