Ensuring consistent performance across diverse populations and incorporating fairness into machine learning models are crucial for advancing medical image diagnostics and promoting equitable healthcare. However, many databases do not provide protected attributes or contain unbalanced representations of demographic groups, complicating the evaluation of model performance across different demographics and the application of bias mitigation techniques that rely on these attributes. This study aims to investigate the effectiveness of using the backbone of Foundation Models as an embedding extractor for creating groups that represent protected attributes, such as gender and age. We propose utilizing these groups in different stages of bias mitigation, including pre-processing, in-processing, and evaluation. Using databases in and out-of-distribution scenarios, it is possible to identify that the method can create groups that represent gender in both databases and reduce in 4.44% the difference between the gender attribute in-distribution and 6.16% in out-of-distribution. However, the model lacks robustness in handling age attributes, underscoring the need for more fundamentally fair and robust Foundation models. These findings suggest a role in promoting fairness assessment in scenarios where we lack knowledge of attributes, contributing to the development of more equitable medical diagnostics.

Text classification, a classic task in natural language processing (NLP), involves assigning predefined categories to textual data and is crucial for applications ranging from sentiment analysis to spam detection. This thesis advances text classification by harnessing the intrinsic knowledge of Pretrained Language Models (PLMs) to address three challenging scenarios: distractor selection for multiple-choice cloze questions, improving robustness for prompt-based zero-shot text classification, and demonstration selection for retrieval-based in-context learning. Firstly, we focus on selecting distractors for multiple-choice cloze questions, ensuring that they are misleading yet incorrect. We assess the relationship between human experts' annotations (accept/reject) and various features, including context-free features (e.g., word frequency) and context-sensitive features (e.g., conditional probabilities of fillin-the-blank words). We utilize pretrained embeddings and follow annotation instructions for context-free feature design, and we find that using contextualized word representations from PLMs as features drastically improves performance over traditional feature-based models, even rivaling human performance (Chapter 3).

In an era where digital text is proliferating at an unprecedented rate, efficient summarization tools are becoming indispensable. While Large Language Models (LLMs) have been successfully applied in various NLP tasks, their role in extractive text summarization remains underexplored. This paper introduces EYEGLAXS (Easy Yet Efficient larGe LAnguage model for eXtractive Summarization), a framework that leverages LLMs, specifically LLAMA2-7B and ChatGLM2-6B, for extractive summarization of lengthy text documents. Instead of abstractive methods, which often suffer from issues like factual inaccuracies and hallucinations, EYEGLAXS focuses on extractive summarization to ensure factual and grammatical integrity. Utilizing state-of-the-art techniques such as Flash Attention and Parameter-Efficient Fine-Tuning (PEFT), EYEGLAXS addresses the computational and resource challenges typically associated with LLMs. The system sets new performance benchmarks on well-known datasets like PubMed and ArXiv. Furthermore, we extend our research through additional analyses that explore the adaptability of LLMs in handling different sequence lengths and their efficiency in training on smaller datasets. These contributions not only set a new standard in the field but also open up promising avenues for future research in extractive text summarization.

It was June 12, Lover's Day in Brazil. Leo Nunes, 24, had spent a few days talking to someone he met on Hornet, a popular gay dating app, before arranging their first encounter in Sao Paulo's middle-class Sacoma neighborhood. A security camera captured the moment that two men on a motorcycle showed up in the alley where he was waiting, grabbed his phone and shot him dead. The Nunes family, who shared details of the investigation with Reuters, said one suspect had been arrested. Sao Paulo police said they are investigating the shooting as a robbery resulting in a homicide, but did not provide further information or confirm if there had been an arrest.

Kyiv, Ukraine – Russia's aerial attack on Ukraine was colossal. Moving in waves from several directions and at different speeds and heights, 127 missiles and 109 drones attacked 15 of Ukraine's 24 regions. The attack is being seen in Ukraine as Russian President Vladimir Putin's revenge for Kyiv's daring incursion into the western Russian region of Kursk that began in early August and has resulted in the apparent takeover of more than 1,000sq kilometres (386sq miles). "He is a vindictive person, he got offended," General Lieutenant Ihor Romanenko, ex-deputy head of the General Staff of Ukraine's Armed Forces, told Al Jazeera. The attack began in predawn darkness on Monday as buzzing swarms of explosives-laden heavy drones took off from the Azov Sea town of Yeisk in southwestern Russia.

The Colombian Nobel laureate Gabriel García Márquez, who spent some of his happiest years chronicling life in Caracas, once declared journalism "the best job in the world". Not so if you are reporting on today's Venezuela, where journalists are feeling the heat as the South American country lurches towards full-blown dictatorship under President Nicolás Maduro. In the four weeks since Venezuela's disputed election, local journalists have come up with a distinctly 21st-century tactic to avoid being arrested for reporting on 21st-century socialism: using artificial intelligence avatars to report all the news Maduro's regime deems unfit to print. In daily broadcasts, the AI-created newsreaders have been telling the world about the president's post-election crackdown on opponents, activists and the media, without putting the reporters behind the stories at risk. Carlos Eduardo Huertas, the director of Connectas, the Colombia-based journalism platform coordinating the initiative, said far from being a gimmick, the use of AI was a response to "the persecution and the growing repression that our colleagues are suffering in Venezuela, where the uncertainty over the safety of doing their job … grows by the minute".

Populism is a concept that is often used but notoriously difficult to measure. Common qualitative measurements like holistic grading or content analysis require great amounts of time and labour, making it difficult to quickly scope out which politicians should be classified as populist and which should not, while quantitative methods show mixed results when it comes to classifying populist rhetoric. In this paper, we develop a pipeline to train and validate an automated classification model to estimate the use of populist language. We train models based on sentences that were identified as populist and pluralist in 300 US governors' speeches from 2010 to 2018 and in 45 speeches of presidential candidates in 2016. We find that these models classify most speeches correctly, including 84% of governor speeches and 89% of presidential speeches. These results extend to different time periods (with 92% accuracy on more recent American governors), different amounts of data (with as few as 70 training sentences per category achieving similar results), and when classifying politicians instead of individual speeches. This pipeline is thus an effective tool that can optimise the systematic and swift classification of the use of populist language in politicians' speeches.

Advanced AI systems capable of generating humanlike text and multimodal content are now widely available. In this paper, we discuss the impacts that generative artificial intelligence may have on democratic processes. We consider the consequences of AI for citizens' ability to make informed choices about political representatives and issues (epistemic impacts). We ask how AI might be used to destabilise or support democratic mechanisms like elections (material impacts). Finally, we discuss whether AI will strengthen or weaken democratic principles (foundational impacts). It is widely acknowledged that new AI systems could pose significant challenges for democracy. However, it has also been argued that generative AI offers new opportunities to educate and learn from citizens, strengthen public discourse, help people find common ground, and to reimagine how democracies might work better.

Computational pathology (CPath) has significantly advanced the clinical practice of pathology. Despite the progress made, Multiple Instance Learning (MIL), a promising paradigm within CPath, continues to face challenges, particularly related to incomplete information utilization. Existing frameworks, such as those based on Convolutional Neural Networks (CNNs), attention, and selective scan space state sequential model (SSM), lack sufficient flexibility and scalability in fusing diverse features, and cannot effectively fuse diverse features. Additionally, current approaches do not adequately exploit order-related and order-independent features, resulting in suboptimal utilization of sequence information. To address these limitations, we propose a novel MIL framework called Mamba2MIL. Our framework utilizes the state space duality model (SSD) to model long sequences of patches of whole slide images (WSIs), which, combined with weighted feature selection, supports the fusion processing of more branching features and can be extended according to specific application needs. Moreover, we introduce a sequence transformation method tailored to varying WSI sizes, which enhances sequence-independent features while preserving local sequence information, thereby improving sequence information utilization. Extensive experiments demonstrate that Mamba2MIL surpasses state-of-the-art MIL methods. We conducted extensive experiments across multiple datasets, achieving improvements in nearly all performance metrics. Specifically, on the NSCLC dataset, Mamba2MIL achieves a binary tumor classification AUC of 0.9533 and an accuracy of 0.8794. On the BRACS dataset, it achieves a multiclass classification AUC of 0.7986 and an accuracy of 0.4981. The code is available at https://github.com/YuqiZhang-Buaa/Mamba2MIL.

Numerous methods and pipelines have recently emerged for the automatic extraction of knowledge graphs from documents such as scientific publications and patents. However, adapting these methods to incorporate alternative text sources like micro-blogging posts and news has proven challenging as they struggle to model open-domain entities and relations, typically found in these sources. In this paper, we propose an enhanced information extraction pipeline tailored to the extraction of a knowledge graph comprising open-domain entities from micro-blogging posts on social media platforms. Our pipeline leverages dependency parsing and classifies entity relations in an unsupervised manner through hierarchical clustering over word embeddings. We provide a use case on extracting semantic triples from a corpus of 100 thousand tweets about digital transformation and publicly release the generated knowledge graph. On the same dataset, we conduct two experimental evaluations, showing that the system produces triples with precision over 95% and outperforms similar pipelines of around 5% in terms of precision, while generating a comparatively higher number of triples.