Paolo Papotti

Matthias Urban, Vu Huy Nguyen, Gabriele Sanmartino et al.

Semantic data systems integrate Large Language Models (LLMs) and Vision-Language Models (VLMs) directly into database query execution, enabling expressive queries on multi-modal data. However, optimizing these queries requires accurate selectivity estimates to determine the most efficient operator execution order. Contemporary systems rely on online sample-based profiling, a process that incurs severe latency overheads and struggles with low-selectivity queries. In this paper, we introduce Semantic Histograms, a novel selectivity estimator for semantic filters on image data that leverages shared embedding spaces to bypass traditional profiling. We realize that all semantic filters are implicit range queries, as they match a range of different images. Some filter predicates are more general, yielding a wide range, while others are more specific, yielding a smaller range. To address the challenge of implicit ranges, we propose two approaches to estimate the queries' specificity, with an ensemble of the two performing best. The evaluation shows that Semantic Histograms can reduce the end-to-end runtime overhead of query optimization and execution by up to 86%.

Mohammed Saeed, Nicolas Traub, Maelle Nicolas et al.

Fact-checking is one of the effective solutions in fighting online misinformation. However, traditional fact-checking is a process requiring scarce expert human resources, and thus does not scale well on social media because of the continuous flow of new content to be checked. Methods based on crowdsourcing have been proposed to tackle this challenge, as they can scale with a smaller cost, but, while they have shown to be feasible, have always been studied in controlled environments. In this work, we study the first large-scale effort of crowdsourced fact-checking deployed in practice, started by Twitter with the Birdwatch program. Our analysis shows that crowdsourcing may be an effective fact-checking strategy in some settings, even comparable to results obtained by human experts, but does not lead to consistent, actionable results in others. We processed 11.9k tweets verified by the Birdwatch program and report empirical evidence of i) differences in how the crowd and experts select content to be fact-checked, ii) how the crowd and the experts retrieve different resources to fact-check, and iii) the edge the crowd shows in fact-checking scalability and efficiency as compared to expert checkers.

Mohammed Saeed, Nicola De Cao, Paolo Papotti

In many use-cases, information is stored in text but not available in structured data. However, extracting data from natural language text to precisely fit a schema, and thus enable querying, is a challenging task. With the rise of pre-trained Large Language Models (LLMs), there is now an effective solution to store and use information extracted from massive corpora of text documents. Thus, we envision the use of SQL queries to cover a broad range of data that is not captured by traditional databases by tapping the information in LLMs. To ground this vision, we present Galois, a prototype based on a traditional database architecture, but with new physical operators for querying the underlying LLM. The main idea is to execute some operators of the the query plan with prompts that retrieve data from the LLM. For a large class of SQL queries, querying LLMs returns well structured relations, with encouraging qualitative results. Preliminary experimental results make pre-trained LLMs a promising addition to the field of database systems, introducing a new direction for hybrid query processing. However, we pinpoint several research challenges that must be addressed to build a DBMS that exploits LLMs. While some of these challenges necessitate integrating concepts from the NLP literature, others offer novel research avenues for the DB community.

Kensuke Mitsuzawa, Motonobu Kanagawa, Stefano Bortoli et al.

We study two-sample variable selection: identifying variables that discriminate between the distributions of two sets of data vectors. Such variables help scientists understand the mechanisms behind dataset discrepancies. Although domain-specific methods exist (e.g., in medical imaging, genetics, and computational social science), a general framework remains underdeveloped. We make two separate contributions. (i) We introduce a mathematical notion of the discriminating set of variables: the largest subset containing no variables whose marginals are identical across the two distributions and independent of the remaining variables. We prove this set is uniquely defined and establish further properties, making it a suitable ground truth for theory and evaluation. (ii) We propose two methods for two-sample variable selection that assign weights to variables and optimise them to maximise the power of a kernel two-sample test while enforcing sparsity to downweight redundant variables. To select the regularisation parameter - unknown in practice, as it controls the number of selected variables - we develop two data-driven procedures to balance recall and precision. Synthetic experiments show improved performance over baselines, and we illustrate the approach on two applications using datasets from water-pipe and traffic networks.

Giulio Corallo, Paolo Papotti

Recent large language model applications, such as Retrieval-Augmented Generation and chatbots, have led to an increased need to process longer input contexts. However, this requirement is hampered by inherent limitations. Architecturally, models are constrained by a context window defined during training. Additionally, processing extensive texts requires substantial GPU memory. We propose a novel approach, Finch, to compress the input context by leveraging the pre-trained model weights of the self-attention. Given a prompt and a long text, Finch iteratively identifies the most relevant Key (K) and Value (V) pairs over chunks of the text conditioned on the prompt. Only such pairs are stored in the KV cache, which, within the space constrained by the context window, ultimately contains a compressed version of the long text. Our proposal enables models to consume large inputs even with high compression (up to 93x) while preserving semantic integrity without the need for fine-tuning.

Giulio Corallo, Paolo Papotti

Retrieval Augmented Generation faces a trade-off: concatenating documents in a long prompt enables multi-document reasoning but creates prefill bottlenecks, while encoding document KV caches separately offers speed but breaks cross-document interaction. We propose Parallel Context-of-Experts Decoding (Pced), a training-free framework that shifts evidence aggregation from the attention mechanism to the decoding. Pced treats retrieved documents as isolated "experts", synchronizing their predictions via a novel retrieval-aware contrastive decoding rule that weighs expert logits against the model prior. This approach recovers cross-document reasoning capabilities without constructing a shared attention across documents.

Francesco Dente, Dario Satriani, Paolo Papotti

Large Language Model (LLM) agents demonstrate strong performance in autonomous code generation under loose specifications. However, production-grade software requires strict adherence to structural constraints, such as architectural patterns, databases, and object-relational mappings. Existing benchmarks often overlook these non-functional requirements, rewarding functionally correct but structurally arbitrary solutions. We present a systematic study evaluating how well agents handle structural constraints in multi-file backend generation. By fixing a unified API contract across 80 greenfield generation tasks and 20 feature-implementation tasks spanning eight web frameworks, we isolate the effect of structural complexity using a dual evaluation with end-to-end behavioral tests and static verifiers. Our findings reveal a phenomenon of constraint decay: as structural requirements accumulate, agent performance exhibits a substantial decline. Capable configurations lose 30 points on average in assertion pass rates from baseline to fully specified tasks, while some weaker configurations approach zero. Framework sensitivity analysis exposes significant performance disparities: agents succeed in minimal, explicit frameworks (e.g., Flask) but perform substantially worse on average in convention-heavy environments (e.g., FastAPI, Django). Finally, error analysis identifies data-layer defects (e.g., incorrect query composition and ORM runtime violations) as the leading root causes. This work highlights that jointly satisfying functional and structural requirements remains a key open challenge for coding agents.

Giuseppe A. Orlando, Paolo Papotti, Maria A. Zuluaga et al.

Vision-language models (VLMs) have shown potential for automated radiology report generation, yet existing approaches rely on global embedding compression of volumetric data, often leading to hallucinated findings and limited anatomical grounding in 3D CT imaging. We introduce MedScribe, a hypothesis-driven framework that reformulates report generation as an iterative evidence acquisition process rather than a single-pass encoding task. MedScribe models reporting as a sequential decision process in which a large language model dynamically invokes pathology-specific diagnostic tools to extract localized volumetric features. These structured features are used to query a multidimensional retrieval space aligned with pathology-specific textual evidence. By explicitly accumulating quantitative evidence prior to synthesis, the framework enforces fine-grained grounding and reduces unsupported claims. Without task-specific fine-tuning, MedScribe improves clinical accuracy, factual consistency, and interpretability on CT-RATE and RadChestCT compared to state-of-the-art 2D and 3D VLMs, demonstrating the value of hypothesis-driven reasoning for reliable medical image reporting.

Simone Papicchio, Simone Rossi, Luca Cagliero et al.

Large Language Models (LLMs) have shown impressive capabilities in transforming natural language questions about relational databases into SQL queries. Despite recent improvements, small LLMs struggle to handle questions involving multiple tables and complex SQL patterns under a Zero-Shot Learning (ZSL) setting. Supervised Fine-Tuning (SFT) partially compensates for the knowledge deficits in pretrained models but falls short while dealing with queries involving multi-hop reasoning. To bridge this gap, different LLM training strategies to reinforce reasoning capabilities have been proposed, ranging from leveraging a thinking process within ZSL, including reasoning traces in SFT, or adopt Reinforcement Learning (RL) strategies. However, the influence of reasoning on Text2SQL performance is still largely unexplored. This paper investigates to what extent LLM reasoning capabilities influence their Text2SQL performance on four benchmark datasets. To this end, it considers the following LLM settings: (1) ZSL, including general-purpose reasoning or not; (2) SFT, with and without task-specific reasoning traces; (3) RL, exploring the use of different rewarding functions, both the established EXecution accuracy (EX) and a mix with fine-grained ones that also account the precision, recall, and cardinality of partially correct answers; (4) SFT+RL, i.e, a two-stage approach that combines SFT and RL. The results show that general-purpose reasoning under ZSL proves to be ineffective in tackling complex Text2SQL cases. Small LLMs benefit from SFT with reasoning much more than larger ones. RL is generally beneficial across all tested models and datasets. The use of the fine-grained metrics turns out to be the most effective RL strategy. Thanks to RL and the novel text2SQL rewards, the 7B Qwen-Coder-2.5 model performs on par with 400+ Billion ones (including gpt-4o) on the Bird dataset.

Riccardo Cappuzzo, Aimee Coelho, Felix Lefebvre et al.

Machine-learning from a disparate set of tables, a data lake, requires assembling features by merging and aggregating tables. Data discovery can extend autoML to data tables by automating these steps. We present an in-depth analysis of such automated table augmentation for machine learning tasks, analyzing different methods for the three main steps: retrieving joinable tables, merging information, and predicting with the resultant table. We use two data lakes: Open Data US, a well-referenced real data lake, and a novel semi-synthetic dataset, YADL (Yet Another Data Lake), which we developed as a tool for benchmarking this data discovery task. Systematic exploration on both lakes outlines 1) the importance of accurately retrieving candidate tables to join, 2) the efficiency of simple merging methods, and 3) the resilience of tree-based learners to noisy conditions. Our experimental environment is easily reproducible and based on open data, to foster more research on feature engineering, autoML, and learning in data lakes.

Alberto Sánchez Pérez, Alaa Boukhary, Paolo Papotti et al.

Generating insightful and actionable information from databases is critical in data analysis. This paper introduces a novel approach using Large Language Models (LLMs) to automatically generate textual insights. Given a multi-table database as input, our method leverages LLMs to produce concise, text-based insights that reflect interesting patterns in the tables. Our framework includes a Hypothesis Generator to formulate domain-relevant questions, a Query Agent to answer such questions by generating SQL queries against a database, and a Summarization module to verbalize the insights. The insights are evaluated for both correctness and subjective insightfulness using a hybrid model of human judgment and automated metrics. Experimental results on public and enterprise databases demonstrate that our approach generates more insightful insights than other approaches while maintaining correctness.

Giulio Corallo, Orion Weller, Fabio Petroni et al.

Incorporating external knowledge in large language models (LLMs) enhances their utility across diverse applications, but existing methods have trade-offs. Retrieval-Augmented Generation (RAG) fetches evidence via similarity search, but key information may fall outside top ranked results. Long-context models can process multiple documents but are computationally expensive and limited by context window size. Inspired by students condensing study material for open-book exams, we propose task-aware key-value (KV) cache compression, which compresses external knowledge in a zero- or few-shot setup. This enables LLMs to reason efficiently over a compacted representation of all relevant information. Experiments show our approach outperforms both RAG and task-agnostic compression methods. On LongBench v2, it improves accuracy by up to 7 absolute points over RAG with a 30x compression rate, while reducing inference latency from 0.43s to 0.16s. A synthetic dataset highlights that RAG performs well when sparse evidence suffices, whereas task-aware compression is superior for broad knowledge tasks.

Francesco Dente, Fabiano Dalpiaz, Paolo Papotti

Large language models (LLMs) can be employed for automating the generation of software requirements from natural language inputs such as the transcripts of elicitation interviews. However, evaluating whether those derived requirements faithfully reflect the stakeholders' needs remains a largely manual task. We introduce Text2Stories, a task and metrics for text-to-story alignment that allow quantifying the extent to which requirements (in the form of user stories) match the actual needs expressed by the elicitation session participants. Given an interview transcript and a set of user stories, our metric quantifies (i) correctness: the proportion of stories supported by the transcript, and (ii) completeness: the proportion of transcript supported by at least one story. We segment the transcript into text chunks and instantiate the alignment as a matching problem between chunks and stories. Experiments over four datasets show that an LLM-based matcher achieves 0.86 macro-F1 on held-out annotations, while embedding models alone remain behind but enable effective blocking. Finally, we show how our metrics enable the comparison across sets of stories (e.g., human vs. generated), positioning Text2Stories as a scalable, source-faithful complement to existing user-story quality criteria.

Azza Abouzied, Firoj Alam, Raian Ali et al.

Misinformation and disinformation pose significant risks globally, with the Arab region facing unique vulnerabilities due to geopolitical instabilities, linguistic diversity, and cultural nuances. We explore these challenges through the key facets of combating misinformation: detection, tracking, mitigation and community-engagement. We shed light on how connecting with grass-roots fact-checking organizations, understanding cultural norms, promoting social correction, and creating strong collaborative information networks can create opportunities for a more resilient information ecosystem in the Arab world.

Dario Satriani, Enzo Veltri, Donatello Santoro et al.

Factuality in Large Language Models (LLMs) is a persistent challenge. Current benchmarks often assess short factual answers, overlooking the critical ability to generate structured, multi-record tabular outputs from parametric knowledge. We demonstrate that this relational fact retrieval is substantially more difficult than isolated point-wise queries, even when individual facts are known to the model, exposing distinct failure modes sensitive to output dimensionality (e.g., number of attributes or records). To systematically evaluate this under-explored capability, we introduce RelationalFactQA, a new benchmark featuring diverse natural language questions (paired with SQL) and gold-standard tabular answers, specifically designed to assess knowledge retrieval in a structured format. RelationalFactQA enables analysis across varying query complexities, output sizes, and data characteristics. Our experiments reveal that even state-of-the-art LLMs struggle significantly, not exceeding 25% factual accuracy in generating relational outputs, with performance notably degrading as output dimensionality increases. These findings underscore critical limitations in current LLMs' ability to synthesize structured factual knowledge and establish RelationalFactQA as a crucial resource for measuring future progress in LLM factuality.

Naser Ahmadi, Hansjorg Sand, Paolo Papotti

Entity resolution is a widely studied problem with several proposals to match records across relations. Matching textual content is a widespread task in many applications, such as question answering and search. While recent methods achieve promising results for these two tasks, there is no clear solution for the more general problem of matching textual content and structured data. We introduce a framework that supports this new task in an unsupervised setting for any pair of corpora, being relational tables or text documents. Our method builds a fine-grained graph over the content of the corpora and derives word embeddings to represent the objects to match in a low dimensional space. The learned representation enables effective and efficient matching at different granularity, from relational tuples to text sentences and paragraphs. Our flexible framework can exploit pre-trained resources, but it does not depends on their existence and achieves better quality performance in matching content when the vocabulary is domain specific. We also introduce optimizations in the graph creation process with an "expand and compress" approach that first identifies new valid relationships across elements, to improve matching, and then prunes nodes and edges, to reduce the graph size. Experiments on real use cases and public datasets show that our framework produces embeddings that outperform word embeddings and fine-tuned language models both in results' quality and in execution times.

Mohammed Saeed, Naser Ahmadi, Preslav Nakov et al.

While pre-trained language models (PLMs) are the go-to solution to tackle many natural language processing problems, they are still very limited in their ability to capture and to use common-sense knowledge. In fact, even if information is available in the form of approximate (soft) logical rules, it is not clear how to transfer it to a PLM in order to improve its performance for deductive reasoning tasks. Here, we aim to bridge this gap by teaching PLMs how to reason with soft Horn rules. We introduce a classification task where, given facts and soft rules, the PLM should return a prediction with a probability for a given hypothesis. We release the first dataset for this task, and we propose a revised loss function that enables the PLM to learn how to predict precise probabilities for the task. Our evaluation results show that the resulting fine-tuned models achieve very high performance, even on logical rules that were unseen at training. Moreover, we demonstrate that logical notions expressed by the rules are transferred to the fine-tuned model, yielding state-of-the-art results on external datasets.

Preslav Nakov, David Corney, Maram Hasanain et al.

The reporting and the analysis of current events around the globe has expanded from professional, editor-lead journalism all the way to citizen journalism. Nowadays, politicians and other key players enjoy direct access to their audiences through social media, bypassing the filters of official cables or traditional media. However, the multiple advantages of free speech and direct communication are dimmed by the misuse of media to spread inaccurate or misleading claims. These phenomena have led to the modern incarnation of the fact-checker -- a professional whose main aim is to examine claims using available evidence and to assess their veracity. As in other text forensics tasks, the amount of information available makes the work of the fact-checker more difficult. With this in mind, starting from the perspective of the professional fact-checker, we survey the available intelligent technologies that can support the human expert in the different steps of her fact-checking endeavor. These include identifying claims worth fact-checking, detecting relevant previously fact-checked claims, retrieving relevant evidence to fact-check a claim, and actually verifying a claim. In each case, we pay attention to the challenges in future work and the potential impact on real-world fact-checking.

Graziano Mita, Paolo Papotti, Maurizio Filippone et al.

We present a novel method - LIBRE - to learn an interpretable classifier, which materializes as a set of Boolean rules. LIBRE uses an ensemble of bottom-up weak learners operating on a random subset of features, which allows for the learning of rules that generalize well on unseen data even in imbalanced settings. Weak learners are combined with a simple union so that the final ensemble is also interpretable. Experimental results indicate that LIBRE efficiently strikes the right balance between prediction accuracy, which is competitive with black box methods, and interpretability, which is often superior to alternative methods from the literature.

Riccardo Cappuzzo, Paolo Papotti, Saravanan Thirumuruganathan

Deep learning based techniques have been recently used with promising results for data integration problems. Some methods directly use pre-trained embeddings that were trained on a large corpus such as Wikipedia. However, they may not always be an appropriate choice for enterprise datasets with custom vocabulary. Other methods adapt techniques from natural language processing to obtain embeddings for the enterprise's relational data. However, this approach blindly treats a tuple as a sentence, thus losing a large amount of contextual information present in the tuple. We propose algorithms for obtaining local embeddings that are effective for data integration tasks on relational databases. We make four major contributions. First, we describe a compact graph-based representation that allows the specification of a rich set of relationships inherent in the relational world. Second, we propose how to derive sentences from such a graph that effectively "describe" the similarity across elements (tokens, attributes, rows) in the two datasets. The embeddings are learned based on such sentences. Third, we propose effective optimization to improve the quality of the learned embeddings and the performance of integration tasks. Finally, we propose a diverse collection of criteria to evaluate relational embeddings and perform an extensive set of experiments validating them against multiple baseline methods. Our experiments show that our framework, EmbDI, produces meaningful results for data integration tasks such as schema matching and entity resolution both in supervised and unsupervised settings.

Naser Ahmadi, Joohyung Lee, Paolo Papotti et al.

One challenge in fact checking is the ability to improve the transparency of the decision. We present a fact checking method that uses reference information in knowledge graphs (KGs) to assess claims and explain its decisions. KGs contain a formal representation of knowledge with semantic descriptions of entities and their relationships. We exploit such rich semantics to produce interpretable explanations for the fact checking output. As information in a KG is inevitably incomplete, we rely on logical rule discovery and on Web text mining to gather the evidence to assess a given claim. Uncertain rules and facts are turned into logical programs and the checking task is modeled as an inference problem in a probabilistic extension of answer set programs. Experiments show that the probabilistic inference enables the efficient labeling of claims with interpretable explanations, and the quality of the results is higher than state of the art baselines.

Fnu Suya, Yuan Tian, David Evans et al.

We study black-box attacks on machine learning classifiers where each query to the model incurs some cost or risk of detection to the adversary. We focus explicitly on minimizing the number of queries as a major objective. Specifically, we consider the problem of attacking machine learning classifiers subject to a budget of feature modification cost while minimizing the number of queries, where each query returns only a class and confidence score. We describe an approach that uses Bayesian optimization to minimize the number of queries, and find that the number of queries can be reduced to approximately one tenth of the number needed through a random strategy for scenarios where the feature modification cost budget is low.