Identifying social isolation themes in NVDRS text narratives using topic modeling and text-classification methods

Social isolation and loneliness, which have been increasing in recent years strongly contribute toward suicide rates. Although social isolation and loneliness are not currently recorded within the US National Violent Death Reporting System's (NVDRS) structured variables, natural language processing (NLP) techniques can be used to identify these constructs in law enforcement and coroner medical examiner narratives. Using topic modeling to generate lexicon development and supervised learning classifiers, we developed high-quality classifiers (average F1: .86, accuracy: .82). Evaluating over 300,000 suicides from 2002 to 2020, we identified 1,198 mentioning chronic social isolation. Decedents had higher odds of chronic social isolation classification if they were men (OR = 1.44; CI: 1.24, 1.69, p<.0001), gay (OR = 3.68; 1.97, 6.33, p<.0001), or were divorced (OR = 3.34; 2.68, 4.19, p<.0001). We found significant predictors for other social isolation topics of recent or impending divorce, child custody loss, eviction or recent move, and break-up. Our methods can improve surveillance and prevention of social isolation and loneliness in the United States.

NLPnorth @ TalentCLEF 2025: Comparing Discriminative, Contrastive, and Prompt-Based Methods for Job Title and Skill Matching

Matching job titles is a highly relevant task in the computational job market domain, as it improves e.g., automatic candidate matching, career path prediction, and job market analysis. Furthermore, aligning job titles to job skills can be considered an extension to this task, with similar relevance for the same downstream tasks. In this report, we outline NLPnorth's submission to TalentCLEF 2025, which includes both of these tasks: Multilingual Job Title Matching, and Job Title-Based Skill Prediction. For both tasks we compare (fine-tuned) classification-based, (fine-tuned) contrastive-based, and prompting methods. We observe that for Task A, our prompting approach performs best with an average of 0.492 mean average precision (MAP) on test data, averaged over English, Spanish, and German. For Task B, we obtain an MAP of 0.290 on test data with our fine-tuned classification-based approach. Additionally, we made use of extra data by pulling all the language-specific titles and corresponding \emph{descriptions} from ESCO for each job and skill. Overall, we find that the largest multilingual language models perform best for both tasks. Per the provisional results and only counting the unique teams, the ranking on Task A is 5$^{\text{th}}$/20 and for Task B 3$^{\text{rd}}$/14.

MFTCXplain: A Multilingual Benchmark Dataset for Evaluating the Moral Reasoning of LLMs through Hate Speech Multi-hop Explanation

Ensuring the moral reasoning capabilities of Large Language Models (LLMs) is a growing concern as these systems are used in socially sensitive tasks. Nevertheless, current evaluation benchmarks present two major shortcomings: a lack of annotations that justify moral classifications, which limits transparency and interpretability; and a predominant focus on English, which constrains the assessment of moral reasoning across diverse cultural settings. In this paper, we introduce MFTCXplain, a multilingual benchmark dataset for evaluating the moral reasoning of LLMs via hate speech multi-hop explanation using Moral Foundation Theory (MFT). The dataset comprises 3,000 tweets across Portuguese, Italian, Persian, and English, annotated with binary hate speech labels, moral categories, and text span-level rationales. Empirical results highlight a misalignment between LLM outputs and human annotations in moral reasoning tasks. While LLMs perform well in hate speech detection (F1 up to 0.836), their ability to predict moral sentiments is notably weak (F1 < 0.35). Furthermore, rationale alignment remains limited mainly in underrepresented languages. These findings show the limited capacity of current LLMs to internalize and reflect human moral reasoning.

Medical Argument Mining: Exploitation of Scarce Data Using NLI Systems

This work presents an Argument Mining process that extracts argumentative entities from clinical texts and identifies their relationships using token classification and Natural Language Inference techniques. Compared to straightforward methods like text classification, this methodology demonstrates superior performance in data-scarce settings. By assessing the effectiveness of these methods in identifying argumentative structures that support or refute possible diagnoses, this research lays the groundwork for future tools that can provide evidence-based justifications for machine-generated clinical conclusions.

Generative or Discriminative? Revisiting Text Classification in the Era of Transformers

The comparison between discriminative and generative classifiers has intrigued researchers since Efron's seminal analysis of logistic regression versus discriminant analysis. While early theoretical work established that generative classifiers exhibit lower sample complexity but higher asymptotic error in simple linear settings, these trade-offs remain unexplored in the transformer era. We present the first comprehensive evaluation of modern generative and discriminative architectures - Auto-regressive modeling, Masked Language Modeling, Discrete Diffusion, and Encoders for text classification. Our study reveals that the classical 'two regimes' phenomenon manifests distinctly across different architectures and training paradigms. Beyond accuracy, we analyze sample efficiency, calibration, noise robustness, and ordinality across diverse scenarios. Our findings offer practical guidance for selecting the most suitable modeling approach based on real-world constraints such as latency and data limitations.

Creating User-steerable Projections with Interactive Semantic Mapping

Dimensionality reduction (DR) techniques map high-dimensional data into lower-dimensional spaces. Yet, current DR techniques are not designed to explore semantic structure that is not directly available in the form of variables or class labels. We introduce a novel user-guided projection framework for image and text data that enables customizable, interpretable, data visualizations via zero-shot classification with Multimodal Large Language Models (MLLMs). We enable users to steer projections dynamically via natural-language guiding prompts, to specify high-level semantic relationships of interest to the users which are not explicitly present in the data dimensions. We evaluate our method across several datasets and show that it not only enhances cluster separation, but also transforms DR into an interactive, user-driven process. Our approach bridges the gap between fully automated DR techniques and human-centered data exploration, offering a flexible and adaptive way to tailor projections to specific analytical needs.

Transforming Chatbot Text: A Sequence-to-Sequence Approach

Due to advances in Large Language Models (LLMs) such as ChatGPT, the boundary between human-written text and AI-generated text has become blurred. Nevertheless, recent work has demonstrated that it is possible to reliably detect GPT-generated text. In this paper, we adopt a novel strategy to adversarially transform GPT-generated text using sequence-to-sequence (Seq2Seq) models, with the goal of making the text more human-like. We experiment with the Seq2Seq models T5-small and BART which serve to modify GPT-generated sentences to include linguistic, structural, and semantic components that may be more typical of human-authored text. Experiments show that classification models trained to distinguish GPT-generated text are significantly less accurate when tested on text that has been modified by these Seq2Seq models. However, after retraining classification models on data generated by our Seq2Seq technique, the models are able to distinguish the transformed GPT-generated text from human-generated text with high accuracy. This work adds to the accumulating knowledge of text transformation as a tool for both attack -- in the sense of defeating classification models -- and defense -- in the sense of improved classifiers -- thereby advancing our understanding of AI-generated text.

Flick: Few Labels Text Classification using K-Aware Intermediate Learning in Multi-Task Low-Resource Languages

Training deep learning networks with minimal supervision has gained significant research attention due to its potential to reduce reliance on extensive labelled data. While self-training methods have proven effective in semi-supervised learning, they remain vulnerable to errors from noisy pseudo labels. Moreover, most recent approaches to the few-label classification problem are either designed for resource-rich languages such as English or involve complex cascading models that are prone to overfitting. To address the persistent challenge of few-label text classification in truly low-resource linguistic contexts, where existing methods often struggle with noisy pseudo-labels and domain adaptation, we propose Flick. Unlike prior methods that rely on generic multi-cluster pseudo-labelling or complex cascading architectures, Flick leverages the fundamental insight that distilling high-confidence pseudo-labels from a broader set of initial clusters can dramatically improve pseudo-label quality, particularly for linguistically diverse, low-resource settings. Flick introduces a novel pseudo-label refinement component, a departure from traditional pseudo-labelling strategies by identifying and leveraging top-performing pseudo-label clusters. This component specifically learns to distil highly reliable pseudo-labels from an initial broad set by focusing on single-cluster cohesion and leveraging an adaptive top-k selection mechanism. This targeted refinement process is crucial for mitigating the propagation of errors inherent in low-resource data, allowing for robust fine-tuning of pre-trained language models with only a handful of true labels. We demonstrate Flick's efficacy across 14 diverse datasets, encompassing challenging low-resource languages such as Arabic, Urdu, and Setswana, alongside English, showcasing its superior performance and adaptability.

Rectifying Privacy and Efficacy Measurements in Machine Unlearning: A New Inference Attack Perspective

Machine unlearning focuses on efficiently removing specific data from trained models, addressing privacy and compliance concerns with reasonable costs. Although exact unlearning ensures complete data removal equivalent to retraining, it is impractical for large-scale models, leading to growing interest in inexact unlearning methods. However, the lack of formal guarantees in these methods necessitates the need for robust evaluation frameworks to assess their privacy and effectiveness. In this work, we first identify several key pitfalls of the existing unlearning evaluation frameworks, e.g., focusing on average-case evaluation or targeting random samples for evaluation, incomplete comparisons with the retraining baseline. Then, we propose RULI (Rectified Unlearning Evaluation Framework via Likelihood Inference), a novel framework to address critical gaps in the evaluation of inexact unlearning methods. RULI introduces a dual-objective attack to measure both unlearning efficacy and privacy risks at a per-sample granularity. Our findings reveal significant vulnerabilities in state-of-the-art unlearning methods, where RULI achieves higher attack success rates, exposing privacy risks underestimated by existing methods. Built on a game-based foundation and validated through empirical evaluations on both image and text data (spanning tasks from classification to generation), RULI provides a rigorous, scalable, and fine-grained methodology for evaluating unlearning techniques.

GeistBERT: Breathing Life into German NLP

Advances in transformer-based language models have highlighted the benefits of language-specific pre-training on high-quality corpora. In this context, German NLP stands to gain from updated architectures and modern datasets tailored to the linguistic characteristics of the German language. GeistBERT seeks to improve German language processing by incrementally training on a diverse corpus and optimizing model performance across various NLP tasks. It was pre-trained using fairseq with standard hyperparameters, initialized from GottBERT weights, and trained on a large-scale German corpus using Whole Word Masking (WWM). Based on the pre-trained model, we derived extended-input variants using Nystr\"omformer and Longformer architectures with support for sequences up to 8k tokens. While these long-context models were not evaluated on dedicated long-context benchmarks, they are included in our release. We assessed all models on NER (CoNLL 2003, GermEval 2014) and text classification (GermEval 2018 fine/coarse, 10kGNAD) using $F_1$ score and accuracy. The GeistBERT models achieved strong performance, leading all tasks among the base models and setting a new state-of-the-art (SOTA). Notably, the base models outperformed larger models in several tasks. To support the German NLP research community, we are releasing GeistBERT under the MIT license.