Knowledge Graphs (KGs) store structured factual knowledge by linking entities through relationships, crucial for many applications. These applications depend on the KG's factual accuracy, so verifying facts is essential, yet challenging. Expert manual verification is ideal but impractical on a large scale. Automated methods show promise but are not ready for real-world KGs. Large Language Models (LLMs) offer potential with their semantic understanding and knowledge access, yet their suitability and effectiveness for KG fact validation remain largely unexplored. In this paper, we introduce FactCheck, a benchmark designed to evaluate LLMs for KG fact validation across three key dimensions: (1) LLMs internal knowledge; (2) external evidence via Retrieval-Augmented Generation (RAG); and (3) aggregated knowledge employing a multi-model consensus strategy. We evaluated open-source and commercial LLMs on three diverse real-world KGs. FactCheck also includes a RAG dataset with 2+ million documents tailored for KG fact validation. Additionally, we offer an interactive exploration platform for analyzing verification decisions. The experimental analyses demonstrate that while LLMs yield promising results, they are still not sufficiently stable and reliable to be used in real-world KG validation scenarios. Integrating external evidence through RAG methods yields fluctuating performance, providing inconsistent improvements over more streamlined approaches -- at higher computational costs. Similarly, strategies based on multi-model consensus do not consistently outperform individual models, underscoring the lack of a one-fits-all solution. These findings further emphasize the need for a benchmark like FactCheck to systematically evaluate and drive progress on this difficult yet crucial task.

Veracity is a critical dimension of data quality that directly impacts a wide range of tasks. In entity search scenarios, Knowledge Graphs (KGs) such as DBpedia and Wikidata serve as core resources for accessing factual content. The veracity of these KGs is therefore essential for ensuring the reliability and trustworthiness of retrieved entities -- factors that directly influence user confidence in the search system. However, ensuring the truthfulness of entities remains a major challenge due to the complexities associated with the scale, development, and maintenance of KGs.

In Information Retrieval (IR), ground truth creation is a crucial yet resource-intensive task that relies on human experts to build test collections - essential for training and evaluating retrieval models. Large-scale evaluation campaigns, such as TREC and CLEF, demand significant human effort to produce reliable, high-quality annotations. To ease this process, tailored annotation tools are pivotal to supporting assessors and streamlining their workload. To this end, we introduce Doctron, a web-based, dockerized annotation tool designed to streamline ground truth creation for IR tasks. Doctron enables the annotation of both textual documents and images. It supports annotating textual passages, identifying relationships, tagging and linking entities, evaluating document relevance to a topic with graded labels, and performing object detection. It offers a collaborative environment where teams can work with defined user roles and permissions. The integration of Inter Annotator Agreement (IAA) measures helps to identify inconsistencies between annotators, thereby ensuring the reliability and high quality of the annotated ground truth data.

Automated fact-checking systems often struggle with trustworthiness, as they lack transparency in their reasoning processes and fail to handle relationships in data. This work presents FactCheck, a fact verification system topped by a web platform that shows how Large Language Models (LLMs) can be collectively used to verify facts within Knowledge Graphs (KGs). While the underlying verification engine implements a system that combines Retrieval Augmented Generation (RAG) with an ensemble of LLMs to validate KG facts, the platform focuses on making the results of this complex process as transparent and accessible as possible. Users can explore how different models interpret the same evidence, compare their reasoning patterns, and understand the factors that lead to the final verification result. The platform supports technical users who want to analyze the model behavior and general users who need to verify whether the facts in the dataset are correct.

Knowledge Graphs (KGs) are widely used in data-driven applications and downstream tasks, such as virtual assistants, recommendation systems, and semantic search. The accuracy of KGs directly impacts the reliability of the inferred knowledge and outcomes. Therefore, assessing the accuracy of a KG is essential for ensuring the quality of facts used in these tasks. However, the large size of real-world KGs makes manual triple-by-triple annotation impractical, thereby requiring sampling strategies to provide accuracy estimates with statistical guarantees. The current state-of-the-art approaches rely on Confidence Intervals (CIs), derived from frequentist statistics. While efficient, CIs have notable limitations and can lead to interpretation fallacies. In this paper, we propose to overcome the limitations of CIs by using Credible Intervals (CrIs), which are grounded in Bayesian statistics. These intervals are more suitable for reliable post-data inference, particularly in KG accuracy evaluation. We prove that CrIs offer greater reliability and stronger guarantees than frequentist approaches in this context. Additionally, we introduce aHPD, an adaptive algorithm that is more efficient for real-world KGs and statistically robust, addressing the interpretive challenges of CIs.

Exploratory search on Knowledge Graphs (KGs) arises when a user needs to understand and extract insights from an unfamiliar KG. In these exploratory sessions, the users issue a series of queries to identify relevant portions of the KG that can answer their questions, with each query answer informing the formulation of the next query. Despite the widespread adoption of KGs, the needs of current KG exploration use cases are not well understood. This work presents the “Exploratory Search Workflows” (ESW) collection focusing on real-world exploration sessions of an open-domain KG, Wikidata, conducted by 57 M.Sc. Computer Engineering students in two advanced Graph Database course editions. This resource includes 234 real exploratory workflows, each containing an average of 45 SPARQL queries and reference workflows that serve as gold-standard solutions to the proposed tasks. The ESW collection is also available as an RDF graph and accessible via a public SPARQL endpoint. It allows for analysis of real user sessions, understanding query evolution and complexity, and serves as the first query benchmark for KG management systems for exploratory search.

In this paper, we discuss the potential costs that emerge from using a Knowledge Graph (KG) in entity-oriented search without considering its data veracity. We argue for the need for KG veracity analysis to gain insights and propose a scalable assessment framework. Previous assessments focused on relevance, assuming correct KGs, and overlooking the potential risks of misinformation. Our approach strategically allocates annotation resources, optimizing utility and revealing the significant impact of veracity on entity search and card generation. Contributions include a fresh perspective on entity-oriented search extending beyond the conventional focus on relevance, a scalable assessment framework, exploratory experiments highlighting the impact of veracity on ranking and user experience, as well as outlining associated challenges and opportunities.

Datasets play a central role in scholarly communications. However, scholarly graphs are often incomplete, particularly due to the lack of connections between publications and datasets. Therefore, the importance of dataset recommendation—identifying relevant datasets for a scientific paper, an author, or a textual query—is increasing. Although various methods have been proposed for this task, their reproducibility remains unexplored, making it difficult to compare them with new approaches. We reviewed current recommendation methods for scientific datasets, focusing on the most recent and competitive approaches, including an SVM-based model, a bi-encoder retriever, a method leveraging co-authors and citation network embeddings, and a heterogeneous variational graph autoencoder. These approaches underwent a comprehensive analysis under consistent experimental conditions. Our reproducibility efforts show that three methods can be reproduced, while the graph variational autoencoder is challenging due to unavailable code and test datasets. Hence, we re-implemented this method and performed a component-based analysis to examine its strengths and limitations. Furthermore, our study indicated that three out of four considered methods produce subpar results when applied to real-world data instead of specialized datasets with ad-hoc features.

Data accuracy is a central dimension of data quality, especially when dealing with Knowledge Graphs (KGs). Auditing the accuracy of KGs is essential to make informed decisions in entity-oriented services or applications. However, manually evaluating the accuracy of large-scale KGs is prohibitively expensive, and research is focused on developing efficient sampling techniques for estimating KG accuracy. This work addresses the limitations of current KG accuracy estimation methods, which rely on the Wald method to build confidence intervals, addressing reliability issues such as zero-width and overshooting intervals. Our solution, rooted in the Wilson method and tailored for complex sampling designs, overcomes these limitations and ensures applicability across various evaluation scenarios. We show that the presented methods increase the reliability of accuracy estimates by up to two times when compared to the state-of-the-art while preserving or enhancing efficiency. Additionally, this consistency holds regardless of the KG size or topology.

This work presents CoreKB, a Web platform for searching reliable facts over gene expression-cancer associations Knowledge Base (KB). It provides search capabilities over an RDF graph using natural language queries, structured facets, and autocomplete. CoreKB is designed to be intuitive and easy to use for healthcare professionals, medical researchers, and clinicians. The system offers the user a comprehensive overview of the scientific evidence supporting a medical fact. It provides a quantitative comparison between the possible gene-cancer associations a particular fact can reflect.

Fairness and equity have become central to ranking problems in information access systems, such as search engines, recommender systems, or marketplaces. To date, several types of fair ranking measures have been proposed, including diversity, exposure, and pairwise fairness measures. Out of those, pairwise fairness is a family of metrics whose normative grounding has not been clearly explicated, leading to uncertainty with respect to the construct that is being measured and how it relates to stakeholders' desiderata.

Abstract Data-driven algorithms are studied and deployed in diverse domains to support critical decisions, directly impacting people’s well-being. As a result, a growing community of researchers has been investigating the equity of existing algorithms and proposing novel ones, advancing the understanding of risks and opportunities of automated decision-making for historically disadvantaged populations. Progress in fair machine learning and equitable algorithm design hinges on data, which can be appropriately used only if adequately documented. Unfortunately, the algorithmic fairness community, as a whole, suffers from a collective data documentation debt caused by a lack of information on specific resources (opacity) and scatteredness of available information (sparsity). In this work, we target this data documentation debt by surveying over two hundred datasets employed in algorithmic fairness research, and producing standardized and searchable documentation for each of them. Moreover we rigorously identify the three most popular fairness datasets, namely Adult, COMPAS, and German Credit, for which we compile in-depth documentation. This unifying documentation effort supports multiple contributions. Firstly, we summarize the merits and limitations of Adult, COMPAS, and German Credit, adding to and unifying recent scholarship, calling into question their suitability as general-purpose fairness benchmarks. Secondly, we document hundreds of available alternatives, annotating their domain and supported fairness tasks, along with additional properties of interest for fairness practitioners and researchers, including their format, cardinality, and the sensitive attributes they encode. We summarize this information, zooming in on the tasks, domains, and roles of these resources. Finally, we analyze these datasets from the perspective of five important data curation topics: anonymization, consent, inclusivity, labeling of sensitive attributes, and transparency. We discuss different approaches and levels of attention to these topics, making them tangible, and distill them into a set of best practices for the curation of novel resources.

Digital data is a basic form of research product for which citation, and the generation of credit or recognition for authors, are still not well understood. The notion of data credit has therefore recently emerged as a new measure, defined and based on data citation groundwork. Data credit is a real value representing the importance of data cited by a research entity. We can use credit to annotate data contained in a curated scientific database and then as a proxy of the significance and impact of that data in the research world. It is a method that, together with citations, helps recognize the value of data and its creators. In this paper, we explore the problem of Data Credit Distribution, the process by which credit is distributed to the database parts responsible for producing data being cited by a research entity. We adopt as use case the IUPHAR/BPS Guide to Pharmacology (GtoPdb), a widely-used curated scientific relational database. We focus on Select-Project-Join (SPJ) queries under bag semantics, and we define three distribution strategies based on how-provenance, responsibility, and the Shapley value. Using these distribution strategies, we show how credit can highlight frequently used database areas and how it can be used as a new bibliometric measure for data and their curators. In particular, credit rewards data and authors based on their research impact, not only on the citation count. We also show how these distribution strategies vary in their sensitivity to the role of an input tuple in the generation of the output data and reward input tuples differently.

Abstract LEarning TO Rank (LETOR) algorithms are usually trained on annotated corpora where a single relevance label is assigned to each available document‐topic pair. Within the Cranfield framework, relevance labels result from merging either multiple expertly curated or crowdsourced human assessments. In this paper, we explore how to train LETOR models with relevance judgments distributions (either real or synthetically generated) assigned to document‐topic pairs instead of single‐valued relevance labels. We propose five new probabilistic loss functions to deal with the higher expressive power provided by relevance judgments distributions and show how they can be applied both to neural and gradient boosting machine (GBM) architectures. Moreover, we show how training a LETOR model on a sampled version of the relevance judgments from certain probability distributions can improve its performance when relying either on traditional or probabilistic loss functions. Finally, we validate our hypothesis on real‐world crowdsourced relevance judgments distributions. Overall, we observe that relying on relevance judgments distributions to train different LETOR models can boost their performance and even outperform strong baselines such as LambdaMART on several test collections.

Abstract LEarning TO Rank (LETOR) is a research area in the field of Information Retrieval (IR) where machine learning models are employed to rank a set of items. In the past few years, neural LETOR approaches have become a competitive alternative to traditional ones like LambdaMART. However, neural architectures performance grew proportionally to their complexity and size. This can be an obstacle for their adoption in large-scale search systems where a model size impacts latency and update time. For this reason, we propose an architecture-agnostic approach based on a neural LETOR model to reduce the size of its input by up to 60% without affecting the system performance. This approach also allows to reduce a LETOR model complexity and, therefore, its training and inference time up to 50%.

This paper analyzes two state-of-the-art Neural Information Retrieval (NeuIR) models: the Deep Relevance Matching Model (DRMM) and the Neural Vector Space Model (NVSM). Our contributions include: (i) a reproducibility study of two state-of-the-art supervised and unsupervised NeuIR models, where we present the issues we encountered during their reproducibility; (ii) a performance comparison with other lexical, semantic and state-of-the-art models, showing that traditional lexical models are still highly competitive with DRMM and NVSM; (iii) an application of DRMM and NVSM on collections from heterogeneous search domains and in different languages, which helped us to analyze the cases where DRMM and NVSM can be recommended; (iv) an evaluation of the impact of varying word embedding models on DRMM, showing how relevance-based representations generally outperform semantic-based ones; (v) a topic-by-topic evaluation of the selected NeuIR approaches, comparing their performance to the well-known BM25 lexical model, where we perform an in-depth analysis of the different cases where DRMM and NVSM outperform the BM25 model or fail to do so. We run an extensive experimental evaluation to check if the improvements of NeuIR models, if any, over the selected baselines are statistically significant.

The semantic mismatch between query and document terms—i.e., the semantic gap—is a long-standing problem in Information Retrieval (IR). Two main linguistic features related to the semantic gap that can be exploited to improve retrieval are synonymy and polysemy. Recent works integrate knowledge from curated external resources into the learning process of neural language models to reduce the effect of the semantic gap. However, these knowledge-enhanced language models have been used in IR mostly for re-ranking and not directly for document retrieval. We propose the Semantic-Aware Neural Framework for IR (SAFIR), an unsupervised knowledge-enhanced neural framework explicitly tailored for IR. SAFIR jointly learns word, concept, and document representations from scratch. The learned representations encode both polysemy and synonymy to address the semantic gap. SAFIR can be employed in any domain where external knowledge resources are available. We investigate its application in the medical domain where the semantic gap is prominent and there are many specialized and manually curated knowledge resources. The evaluation on shared test collections for medical literature retrieval shows the effectiveness of SAFIR in terms of retrieving and ranking relevant documents most affected by the semantic gap.

RDF datasets are becoming increasingly useful with the development of knowledge-based web applications. SPARQL is the official structured query language to search and access RDF datasets. Despite its effectiveness, the language is often difficult to use for non-experts because of its syntax and the necessity to know the underlying data structure of the database queries. In this regard, keyword search enables non-expert users to access the data contained in RDF datasets intuitively. This work describes the TSA+VDP keyword search system for effective and efficient keyword search over large RDF datasets. The system is compared with other state-of-the-art methods on different datasets, both real-world and synthetic, using a new evaluation framework that is easily reproducible and sharable.

An increasing amount of information is being published in structured databases and retrieved using queries, raising the question of how query results should be cited. Since there are a large number of possible queries over a database, one strategy is to specify citations to a small set of frequent queries - citation views - and use these to construct citations to other "general" queries. We present three approaches to implementing citation views and describe alternative policies for the joint, alternate and aggregated use of citation views. Extensive experiments using both synthetic and realistic citation views and queries show the trade-offs between the approaches in terms of the time to generate citations, as well as the size of the resulting citation. They also show that the choice of policy has a huge effect both on performance and size, leading to useful guidelines for what policies to use and how to specify citation views.

Information retrieval (IR) systems are the prominent means for searching and accessing huge amounts of unstructured information on the web and elsewhere. They are complex systems, constituted by many different components interacting together, and evaluation is crucial to both tune and improve them. Nevertheless, in the current evaluation methodology, there is still no way to determine how much each component contributes to the overall performances and how the components interact together. This hampers the possibility of a deep understanding of IR system behavior and, in turn, prevents us from designing ahead which components are best suited to work together for a specific search task. In this paper, we move the evaluation methodology one step forward by overcoming these barriers and beginning to devise an “anatomy” of IR systems and their internals. In particular, we propose a methodology based on the General Linear Mixed Model (GLMM) and analysis of variance (ANOVA) to develop statistical models able to isolate system variance and component effects as well as their interaction, by relying on a grid of points (GoP) containing all the combinations of the analyzed components. We apply the proposed methodology to the analysis of two relevant search tasks—news search and web search—by using standard TREC collections. We analyze the basic set of components typically part of an IR system, namely, stop lists, stemmers, and n‐grams, and IR models. In this way, we derive insights about English text retrieval.

Citations are the cornerstone of knowledge propagation and the primary means of assessing the quality of research, as well as directing investments in science. Science is increasingly becoming “data‐intensive,” where large volumes of data are collected and analyzed to discover complex patterns through simulations and experiments, and most scientific reference works have been replaced by online curated data sets. Yet, given a data set, there is no quantitative, consistent, and established way of knowing how it has been used over time, who contributed to its curation, what results have been yielded, or what value it has. The development of a theory and practice of data citation is fundamental for considering data as first‐class research objects with the same relevance and centrality of traditional scientific products. Many works in recent years have discussed data citation from different viewpoints: illustrating why data citation is needed, defining the principles and outlining recommendations for data citation systems, and providing computational methods for addressing specific issues of data citation. The current panorama is many‐faceted and an overall view that brings together diverse aspects of this topic is still missing. Therefore, this paper aims to describe the lay of the land for data citation, both from the theoretical (the why and what) and the practical (the how) angle.

Data citation is an interesting computational challenge, whose solution draws on several well-studied problems in database theory: query answering using views, and provenance. We describe the problem, suggest an approach to its solution, and highlight several open research problems, both practical and theoretical.

The practice of citation is foundational for the propagation of knowledge along with scientific development and it is one of the core aspects on which scholarship and scientific publishing rely. Within the broad context of data citation, we focus on the automatic construction of citations problem for hierarchically structured data. We present the “learning to cite” framework, which enables the automatic construction of human‐ and machine‐readable citations with different levels of coarseness. The main goal is to reduce the human intervention on data to a minimum and to provide a citation system general enough to work on heterogeneous and complex XML data sets. We describe how this framework can be realized by a system for creating citations to single nodes within an XML data set and, as a use case, show how it can be applied in the context of digital archives. We conduct an extensive evaluation of the proposed citation system by analyzing its effectiveness from the correctness and completeness viewpoints, showing that it represents a suitable solution that can be easily employed in real‐world environments and that reduces human intervention on data to a minimum.

We present the innovative visual analytics approach of the VATE system, which eases and makes more effective the experimental evaluation process by introducing the what-if analysis. The what-if analysis is aimed at estimating the possible effects of a modification to an IR system to select the most promising fixes before implementing them, thus saving a considerable amount of effort. VATE builds on an analytical framework which models the behavior of the systems in order to make estimations, and integrates this analytical framework into a visual part which, via proper interaction and animations, receives input and provides feedback to the user.

Topic variance has a greater effect on performances than system variance but it cannot be controlled by system developers who can only try to cope with it. On the other hand, system variance is important on its own, since it is what system developers may affect directly by changing system components and it determines the differences among systems. In this paper, we face the problem of studying system variance in order to better understand how much system components contribute to overall performances. To this end, we propose a methodology based on General Linear Mixed Model (GLMM) to develop statistical models able to isolate system variance, component effects as well as their interaction by relying on a Grid of Points (GoP) containing all the combinations of analysed components. We apply the proposed methodology to the analysis of TREC Ad-hoc data in order to show how it works and discuss some interesting outcomes of this new kind of analysis. Finally, we extend the analysis to different evaluation measures, showing how they impact on the sources of variance.

We present a novel measure for ranking evaluation, called Twist (τ). It is a measure for informational intents, which handles both binary and graded relevance. τ stems from the observation that searching is currently a that searching is currently taken for granted and it is natural for users to assume that search engines are available and work well. As a consequence, users may assume the utility they have in finding relevant documents, which is the focus of traditional measures, as granted. On the contrary, they may feel uneasy when the system returns nonrelevant documents because they are then forced to do additional work to get the desired information, and this causes avoidable effort. The latter is the focus of τ, which evaluates the effectiveness of a system from the point of view of the effort required to the users to retrieve the desired information. We provide a formal definition of τ, a demonstration of its properties, and introduce the notion of effort/gain plots, which complement traditional utility‐based measures. By means of an extensive experimental evaluation, τ is shown to grasp different aspects of system performances, to not require extensive and costly assessments, and to be a robust tool for detecting differences between systems.

Digital Geolinguistic systems encourage collaboration between linguists, historians, archaeologists, ethnographers, as they explore the relationship between language and cultural adaptation and change. In this demo, we propose a Linked Open Data approach for increasing the level of interoperability of geolinguistic applications and the reuse of the data. We present a case study of a geolinguistic project named Atlante Sintattico d'Italia, Syntactic Atlas of Italy (ASIt).

The paper addresses the problem of representing, managing and exchanging hierarchically structured data in the context of Digital Library (DL) systems in order to enhance the access and exchange DL resources on the Web. We propose the NEsted SeTs for Object hieRarchies (NESTOR) framework, which relies on two set data models — the “Nested Set Model (NS-M)” and the “Inverse Nested Set Model (INS-M)” — to enable the representation of hierarchical data structures by means of a proper organization of nested sets. In particular, we show how NESTOR can be effectively exploited to enhance Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) for better access and exchange of hierarchical resources on the Web.