arXiv subjects

Suzan Verberne

Publications and source records attributed to Suzan Verberne.

At least 19 recordsLinked to original sources

arXiv2026

Domain-specific information retrieval

Information Retrieval (IR) is much more than ad hoc web search: a large portion of the information seeking activities takes place in specific domains, often addressing work-related tasks. Examples are lawyers finding relevant prior cases for a legal case and academic researchers studying the background material for a paper. In Chapter 2 we outline the most common domain-specific tasks, datasets and applications. Domain-specific IR has a number of challenges that make set it apart from simple web search tasks: they often deal with long documents, long queries and long sessions with specific user needs. We discuss these in Chapter 3. We also address issues related to bias and fairness of models. In Chapter 4, we describe the models and methods developed for domain-specific tasks and data, addressing the specific challenges, including user aspects related to domain-specific IR tasks, dealing with queries, user models, and user interactions. We show that many of the methodological advances in domain-specific IR are transferable to other domains and even to open-domain web search. We conclude with the main takeaways and propose research directions for the near future.

cs.IR↗

arXiv2026

Personalized to Persuade: The Effects of Contextualization and Warmth on Trust and Reliance in Conversational AI

Contextualization is a form of personalization in which the message is framed in a context meaningful to the recipient, using their interests and background. Personalization is a known persuasive strategy of artificial intelligence (AI) systems in politics and marketing. However, the persuasive effect of contextualization in everyday tasks, where users often lack prior knowledge, remains unclear. We ran a $2\times2$ between-subjects experiment ($N = 407$) in a fictional scenario where participants had no prior opinions or knowledge. We examined how contextualization, combined with conversational warmth, shapes the persuasive power of an AI assistant arguing against expert recommendations and users' reliance on it. Our findings show that the neutral, non-contextualized assistant was the most persuasive. Contextualization and warmth each reduced persuasiveness on their own, but their interaction increased it. Reliance was present in all conditions, and neither cue changed it detectably. Trust predicted persuasion and reliance, yet neither contextualization nor warmth had an indirect effect through trust. Participants with higher self-reported AI literacy relied more on the assistant without trusting it more. Our study extends AI personalization research from political opinions to everyday decisions made without prior knowledge. The results suggest that users are more likely to be persuaded by, and defer to, an AI assistant than an expert, even without personalization. AI persuasion experiments should therefore include untailored assistants and measure decisions; the cues that changed persuasion did not detectably change reliance.

cs.HC↗

arXiv2026

UserArena: Benchmarking Interactive User Simulation for Long-Horizon Shopping Agents

Agentic recommendation pursues shopper goals through sequences of context-dependent decisions. Developing and evaluating such policies requires faithful recommendation world models that capture both the interactive shopping environment and users' behavioral responses over long interactions. Existing resources rarely provide long-horizon, fine-grained real-user trajectories or support trajectory-level evaluation and offline validation against production A/B results. We introduce UserArena, a real-world e-commerce interactive benchmark containing 64,729 session trajectories from 11,010 users, spanning 566,102 items and 13 action types. UserArena pairs these trajectories with an interactive environment and supports offline validation against outcomes from a controlled production A/B test, enabling five-level evaluation from output validity and local decisions to sequential behavior, personalization, and outcomes. To improve simulator fidelity over these long-horizon interactions, we further propose On-Policy Trajectory Distillation (OPTD), which uses trajectory-level intent to provide dense teacher targets on student-generated rollouts. Experiments show that OPTD achieves the strongest overall fidelity among the evaluated simulators under UserArena's evaluation protocol and most closely reproduces the production policy effect in offline simulation.

cs.IR↗

arXiv2026

Computed, not Memorized: Bit-Factorized Identifiers for Efficient Generative Retrieval

Generative retrieval is known for its efficient end-to-end indexing and retrieval. In particular, numeric Document Identifiers (DocIDs) are generally more efficient because they are shorter and generally fixed-length sequences over a small vocabulary, requiring fewer decoding steps during beam search. However, they tend to generalize poorly to unseen documents and must be learned during indexing-stage training. In this paper, we propose a numeric DocID design for generative retrieval that enables better adaptation to unseen documents without retraining. Our DocID design inherits the language model's knowledge: each DocID is a sequence of tokens factorized into bits, and each token is scored as the sum of its bits. Unlike existing numeric DocIDs, whose tokens carry meaning only through embeddings memorized during indexing training, each bit here derives its meaning from the model's representation. Therefore, the DocIDs of unseen documents remain scorable without retraining. Experiments on three benchmarks show that our DocID design substantially improves unseen-document retrieval while maintaining comparable performance on seen documents. It achieves over 20 percentage points higher Hit@10 on unseen documents and up to 7 times lower inference latency than existing numeric DocID designs on NQ320K and TriviaQA, while performing only 3 percentage points below weight-update methods on MS MARCO without additional training. These results show that an appropriate DocID design can heavily reduce retraining costs when adapting generative retrievers to new documents.

cs.IR↗

arXiv2026

Activation Steering with Jungian Cognitive Functions

Controlling combinations of personality-related tendencies is important for adapting large language models to diverse interaction needs. But successful single-trait steering does not establish how to construct interventions for joint targets, or whether existing trait directions adequately represent them. We introduce a target-first framework that directly extracts steering directions from joint personality targets and then examines their decomposition into individual directions. To support this framework, we construct NarrationDB, linking character self-narrations to automatically evaluated joint scores across eight Jungian-inspired functions, a correspondence not directly available in existing Big Five- or MBTI-oriented resources. Experiments on three models demonstrate questionnaire-score controllability across the eight functions and coordinated changes in a tested two-function target. Directly extracted joint directions are better approximated by their corresponding individual directions than by unrelated pairings, but retain substantial residual components. Our work provides a method and resource for investigating joint personality control, with evidence that individual steering directions capture part, but not all, of the structure of joint targets.

cs.CL↗

arXiv2026

Time as Structure: Temporal Dependency Graphs for Verifiable Deadline Computation over Legal Documents

Miss a filing deadline by one day and the claim is barred, however strong the case. Computing that deadline is rarely simple: the period runs from a triggering event, is counted by a statutory convention, and may be suspended by a mandatory conciliation window. We ask whether a language model should answer such questions directly, or read the document and leave the arithmetic to code. We extract dated facts and their dependencies into a temporal dependency graph and compute deadlines from it with a calendar-correct engine. On UK Employment Appeal Tribunal judgments the engine reproduces six of seven timeliness rulings, and matches the judges' own dates to the day. The strongest of four language models, asked the same cases, gets the arithmetic right and the answer wrong: in two of its seven responses its stated verdict contradicts its own working, and across the three GPT models six of twenty-one responses do, each time calling a late claim timely. To test the systems at scale we move the dismissal date across the statutory boundary, generating 427 cases whose answers are computed rather than annotated. On the cases both systems answer, the pipeline is right 90.2% of the time against 61.2% for direct answering. It answers two thirds of the cases, and counting its abstentions as errors it does not exceed direct answering (60.4% against 64.6%). The limit is extraction: on contracts the errors are almost never in the arithmetic, but in choosing which event the period starts from.

cs.CL↗

arXiv2026

Is Memorization Context-Sensitive? Prefix-Based Extraction Beyond Isolated Prefixes

Large language models (LLMs) can expose memorized training sequences under prefix-based extraction: given a prefix from a training example, the model may assign high probability to the original continuation. In deployed systems, however, prefixes are rarely evaluated in isolation. They often appear together with instructions, retrieved documents, or other task-specific context, as in retrieval-augmented generation (RAG). This motivates examining whether contextual conditioning mitigates memorization or merely changes the set of memorized samples that become extractable. We investigate this issue through paired item-level measurements of probabilistic suffix extraction. For each prefix-suffix pair, we score the target suffix under an empty prompt and under retrieved contexts of varying relevance, across three open-weight instruction-tuned models. We find that context does not simply erase memorization. Instead, extractable memorization consists of a context-robust core and a context-sensitive boundary. Many samples that are extractable without context remain extractable under the retrieved context, especially as the prefix length increases. At the same time, context mainly affects marginal samples near the extraction threshold: it suppresses some exposures, but also enables new ones that are missed by prefix-only evaluation. These findings qualify the view that RAG reduces memorization risk. Context can lower aggregate extraction by suppressing boundary cases, yet robustly extractable samples persist, and context-enabled extractability remains security-relevant.

cs.AI↗

arXiv2026

HealthcareNLP: where are we and what is next?

This tutorial focused on Healthcare Domain Applications of NLP, what we have achieved around HealthcareNLP, and the challenges that lie ahead for the future. Existing reviews in this domain either overlook some important tasks, such as synthetic data generation for addressing privacy concerns, or explainable clinical NLP for improved integration and implementation, or fail to mention important methodologies, including retrieval augmented generation and the neural symbolic integration of LLMs and KGs. In light of this, the goal of this tutorial is to provide an introductory overview of the most important sub-areas of a patient- and resource-oriented HealthcareNLP, with three layers of hierarchy: data/resource layer: annotation guidelines, ethical approvals, governance, synthetic data; NLP-Eval layer: NLP tasks such as NER, RE, sentiment analysis, and linking/coding with categorised methods, leading to explainable HealthAI; patients layer: Patient Public Involvement and Engagement (PPIE), health literacy, translation, simplification, and summarisation (also NLP tasks), and shared decision-making support. A hands-on session will be included in the tutorial for the audience to use HealthcareNLP applications. The target audience includes NLP practitioners in the healthcare application domain, NLP researchers who are interested in domain applications, healthcare researchers, and students from NLP fields. The type of tutorial is "Introductory to CL/NLP topics (HealthcareNLP)" and the audience does not need prior knowledge to attend this. Tutorial materials: https://github.com/4dpicture/HealthNLP

cs.CL↗

arXiv2026

Breaking Bureaucracy: Evaluating open-source LLMs for legal document review

In this paper, we evaluate open-source generative LLMs on legal Natural Language Inference (NLI). Legal inspectorial processes take place in specific domains and often deal with confidential data. This creates a need for working with local models that do not require labeled training data. We evaluate our models on the ContractNLI benchmark and two NLI4Wills datasets. We successfully reproduce the baseline for the task (Span NLI BERT) and we evaluate multiple open-source LLMs on the same task. We analyze the invalid rate of the models, and their stability across temperature settings and domains. Among the generative models, Gemma-4 26B performs the best, reaching an accuracy of 81.2%, even outperforming the supervised model on one metric. On accuracy, it is not possible to beat the supervised model with zero-shot approaches. Qwen-3.6 35B performs well on both ContractNLI and additional datasets in the legal wills domain. Our findings indicate that zero-shot, open-source, generative LLMs are a viable alternative for real-world legal NLI when no supervised data is available. Our code is available at https://github.com/fbaratov/contractnli-llms.

cs.CL↗

arXiv2026

Walking the Embedding Space: Datastore Extraction from Multimodal RAG

Multimodal Retrieval-Augmented Generation (MRAG) has emerged as a reliable and cost-effective technique of grounding the generative capabilities of Multimodal Large Language Models (MLLMs) into relevant, up-to-date, external knowledge. Despite presenting several benefits, such as reducing hallucinatory behavior, they also introduce new attack surfaces, including leakage of private information and vulnerabilities against data extraction attacks. In this paper, we introduce $\immrag$, an adaptive and automatic data extraction attack procedure operating in a black box setting against \emph{image-returning} MRAG, a configuration in which the retrieved visual artifact is itself the response. Each query blends an attacker-held shadow image with an image already recovered from the system, and relevance-weighted resampling steers subsequent queries towards regions of the embedding space that still yield novel retrievals. Unlike current extraction attacks that aim to persuade the model towards data leakage by placing a malicious query as a textual prompt, $\immrag$ embeds the malicious instructions inside a user-given input image. We evaluate $\immrag$ on three plausible and distinct real-world scenarios: medical assistant, document-focused helper and general purpose tool. The experiments involve the study of the effectiveness of the attack on multiple CLIP-family retrievers, as well as the impact of various generators. A single 2500-query run reconstructs up to 611 distinct radiology images, 566 document scans and 416 general-purpose images under local-feature correspondence, and reaches up to $5.6\times$ as many distinct datastore items as a non-adaptive baseline. Our results show the urgent need for safeguards specifically designed for multimodal data.

cs.CR↗

arXiv2026

RankSteer: Can Pointwise LLM Rankers Be Calibrated at the Representation Level?

Large language models (LLMs) are strong zero-shot pointwise rankers, but lag behind pairwise and listwise methods. Beyond missing comparative signals, we identify a \textit{calibration gap}: ranking-relevant information encoded in hidden states is not fully captured by the scalar output head. We propose RankSteer, a post-hoc activation-steering framework that calibrates ranking via projection-based interventions along multiple directions at inference time: decision, evidence, and, optionally, role. This is achieved without updating model weights or introducing cross-document comparisons. We instantiate RankSteer on two structurally distinct pointwise variants and observe improvements over their respective baselines on most TREC DL and BEIR datasets across three backbones. This suggests that the calibration gap is a general property of pointwise rankers. Our additional geometric analysis shows that steering improves ranking by concentrating each query's document representations along an existing ranking geometry, offering new insight into how LLMs internally represent and calibrate relevance judgments.

cs.IR↗

arXiv2026

Does online sustainability communication shape public discourse? Insights from six years of tenant-housing provider interactions

Authorities increasingly rely on social media to advance sustainability transitions, infrastructure investment, and service reform. Yet how citizens respond to these digital communications remains poorly understood. Existing approaches rely on aggregate engagement metrics (e.g., likes), providing limited insight into discourse structure and quality. We developed a data-driven, multidimensional framework to analyse how social media communication shapes the content of discourse, focusing on sustainability-related engagement in Dutch public housing. We analysed 792 posts and 3,197 tenant comments from the Facebook pages of 92 housing providers (2018-2023). A machine-learning pipeline classified comments into recurring discourse configurations across three dimensions - communicative intent, sentiment, and semantic relatedness. Multinomial logistic regression estimated the effects of post-design and organisational characteristics on discourse. Tenant comments were significantly more semantically aligned with their corresponding posts than with randomly paired content, indicating that organisational communication structures responses to topics. Six discourse types emerged, with critical and inquiry-driven engagement increasing over time. Post-level features did not significantly explain variation; organisational characteristics dominated. Larger housing associations attracted more substantive responses, while lower-rent organisations received fewer evaluative comments. While applied to housing associations, our methodology provides a scalable approach to analyse online discourse dynamics, quality, and content across organisations and contexts.

cs.CY↗

arXiv2026

Measuring the practice of shared-decision making (OPTION12): An Investigation into Open-sourced Smaller LLMs (OS-sLLMs) for Better Privacy and Sustainability

We present LLM4SDM, the first study of open-source smaller language models (OS-sLLMs) for automated assessment of shared decision making (SDM) using the Observer OPTION12 framework. Unlike previous work that relies on large commercial models and the shorter OPTION5 instrument, our study focuses on privacy-preserving locally deployable models and Dutch melanoma consultation transcripts. Using expert-annotated clinical consultations, we evaluate three general-domain and two medical-domain OS-sLLMs during a development-phase pilot study. Results show that general-domain models outperform medical-domain models, which exhibit substantial hallucination and instruction-following failures. Gemma3:12b achieves the strongest agreement with human annotations (Pearson r=0.51, Spearman \r{ho}=0.59). Item-level and qualitative analyses reveal systematic challenges related to temporal discourse reasoning, conversational role attribution, and evidence grounding. We further introduce a Judge-LLM consensus framework designed to support disagreement resolution among multiple models. Our findings suggest that while current OS-sLLMs cannot replace human annotators, they offer a promising foundation for privacy-preserving human-in-the-loop SDM assessment.

cs.CL↗

arXiv2026

OpenReward: Learning to Reward Long-form Agentic Tasks via Reinforcement Learning

Reward models (RMs) have become essential for aligning large language models (LLMs), serving as scalable proxies for human evaluation in both training and inference. However, existing RMs struggle on knowledge-intensive and long-form tasks, where evaluating correctness requires grounding beyond the model's internal knowledge. This limitation hinders them from reliably discriminating subtle quality differences, especially when external evidence is necessary. To address this, we introduce OpenRM, a tool-augmented long-form reward model that systematically judges open-ended responses by invoking external tools to gather relevant evidence. We train OpenRM with Group Relative Policy Optimization (GRPO) on over 27K synthesized pairwise examples generated through a controllable data synthesis framework. The training objective jointly supervises intermediate tool usage and final outcome accuracy, incentivizing our reward model to learn effective evidence-based judgment strategies. Extensive experiments on three newly-collected datasets and two widely-used benchmarks demonstrate that OpenRM substantially outperforms existing reward modeling approaches. As a further step, we integrate OpenRM into both inference-time response selection and training-time data selection. This yields consistent gains in downstream LLM alignment tasks, highlighting the potential of tool-augmented reward models for scaling reliable long-form evaluation.

cs.CL↗

arXiv2026

ReportLogic: Evaluating Logical Quality in Deep Research Reports

Users increasingly rely on Large Language Models (LLMs) for Deep Research, using them to synthesize diverse sources into structured reports that support understanding and action. In this context, the practical reliability of such reports hinges on logical quality: whether the report's claims and arguments are explicitly supported and can be trusted as a basis for downstream use, rather than merely appearing fluent or informative. However, current evaluation frameworks largely overlook this requirement. To bridge this gap, we introduce ReportLogic, a benchmark that quantifies report-level logical quality through a reader-centric lens of auditability. Specifically, ReportLogic adopts a hierarchical taxonomy that evaluates whether readers can (1) trace an on-topic report structure with a unified analytical arc (Macro-Logic), (2) understand the progression with necessary context (Expositional-Logic), and (3) verify conclusions via explicit claim--support (Structural-Logic). Based on this taxonomy, we construct a human-annotated rubric-guided dataset and train an open-source LogicJudge for scalable evaluation. We further evaluate judge robustness via adversarial attacks, showing that off-the-shelf LLM judges are frequently influenced by superficial cues (e.g., verbosity), and reasoning modes can mask broken support relations. Overall, our results provide actionable guidance for building more robust logic evaluators and improving the logical reliability of LLM-generated reports.

cs.CL↗

arXiv2026

Unifying Search and Recommendation in LLMs via Gradient Multi-Subspace Tuning

Search and recommendation (S&R) are core to online platforms, addressing explicit intent through queries and modeling implicit intent from behaviors, respectively. Their complementary roles motivate a unified modeling paradigm. Early studies to unify S&R adopt shared encoders with task-specific heads, while recent efforts reframe item ranking in both S&R as conditional generation. The latter holds particular promise, enabling end-to-end optimization and leveraging the semantic understanding of LLMs. However, existing methods rely on full fine-tuning, which is computationally expensive and limits scalability. Parameter-efficient fine-tuning (PEFT) offers a more practical alternative but faces two critical challenges in unifying S&R: (1) gradient conflicts across tasks due to divergent optimization objectives, and (2) shifts in user intent understanding caused by overfitting to fine-tuning data, which distort general-domain knowledge and weaken LLM reasoning. To address the above issues, we propose Gradient Multi-Subspace Tuning (GEMS), a novel framework that unifies S&R with LLMs while alleviating gradient conflicts and preserving general-domain knowledge. GEMS introduces (1) \textbf{Multi-Subspace Decomposition}, which disentangles shared and task-specific optimization signals into complementary low-rank subspaces, thereby reducing destructive gradient interference, and (2) \textbf{Null-Space Projection}, which constrains parameter updates to a subspace orthogonal to the general-domain knowledge space, mitigating shifts in user intent understanding. Extensive experiments on benchmark datasets show that GEMS consistently outperforms the state-of-the-art baselines across both search and recommendation tasks, achieving superior effectiveness.

cs.IR↗

arXiv2026

Changing topic bias in biomedical science maps by linking documents through alternative data sources: policy documents, patents, authors, Facebook, and Twitter

Traditional science maps visualize topics by clustering documents within a network, but they are inherently biased toward clustering certain topics over others. If these topics could be chosen, then the science maps could be tailored for different needs. In this paper, we explore the extent to which the topic bias of a science map can be changed by choosing different data sources to build the document network. We analyze this by evaluating the clustering effectiveness of several topic categories over two sources that are traditionally used for the creation of science maps (citations and text similarity) and six non-traditional data sources, which we found favor different kinds of topics: Health issues for Facebook users, biotechnology topics for patent families, government and social issues for policy documents, food topics for Twitter conversations, nursing topics for Twitter users, and geographical entities for document authors (the favoring in this latter source was particularly strong). Our results show that diverse data sources can be used to control topic bias, which opens up the possibility of creating science maps tailored for different needs.

cs.DL↗

arXiv2026

Analyzing Cancer Patients' Experiences with Embedding-based Topic Modeling and LLMs

This study investigates the use of neural topic modeling and LLMs to uncover meaningful themes from patient storytelling data, to offer insights that could contribute to more patient-oriented healthcare practices. We analyze a collection of transcribed interviews with cancer patients (132,722 words in 13 interviews). We first evaluate BERTopic and Top2Vec for individual interview summarization by using similar preprocessing, chunking, and clustering configurations to ensure a fair comparison on Keyword Extraction. LLMs (GPT4) are then used for the next step topic labeling. Their outputs for a single interview (I0) are rated through a small-scale human evaluation, focusing on {coherence}, {clarity}, and {relevance}. Based on the preliminary results and evaluation, BERTopic shows stronger performance and is selected for further experimentation using three {clinically oriented embedding} models. We then analyzed the full interview collection with the best model setting. Results show that domain-specific embeddings improved topic \textit{precision} and \textit{interpretability}, with BioClinicalBERT producing the most consistent results across transcripts. The global analysis of the full dataset of 13 interviews, using the BioClinicalBERT embedding model, reveals the most dominant topics throughout all 13 interviews, namely ``Coordination and Communication in Cancer Care Management" and ``Patient Decision-Making in Cancer Treatment Journey''. Although the interviews are machine translations from Dutch to English, and clinical professionals are not involved in this evaluation, the findings suggest that neural topic modeling, particularly BERTopic, can help provide useful feedback to clinicians from patient interviews. This pipeline could support more efficient document navigation and strengthen the role of patients' voices in healthcare workflows.

cs.CL↗