The growing demand for efficient knowledge graph (KG) enrichment leveraging external corpora has intensified interest in relation extraction (RE), particularly under low-supervision settings. To address the need for adaptable and noise-resilient RE solutions that integrate seamlessly with pre-trained large language models (PLMs), we introduce SCoRE, a modular and cost-effective sentence-level RE system. SCoRE enables easy PLM switching, requires no finetuning, and adapts smoothly to diverse corpora and KGs. By combining supervised contrastive learning with a Bayesian k-Nearest Neighbors (kNN) classifier for multi-label classification, it delivers robust performance despite the noisy annotations of distantly supervised corpora. To improve RE evaluation, we propose two novel metrics: Correlatin Structure Distance (CSD), measuring the alignment between learned relational patterns and KG structures, and Precision at R (P@R), assessing utility as a recommender system. We also release Wiki20d, a benchmark dataset replicating real-world RE conditions where only KG-derived annotations are available. Experiments on five benchmarks demonstrate that SCoRE matches or slightly surpasses state-of-the-art methods (average gains of +3.2 in micro-F1 and +5.9 in macro-F1 against fully reproducible baselines), while reducing the training burden by more than an order of magnitude ( ≈ 99 % lower energy consumption in kWh) Further analyses reveal that increasing model complexity, as seen in prior work, degrades performance, highlighting the advantages of SCoRE’s minimal design. Combining efficiency, modularity, and scalability, SCoRE stands as an optimal choice for real-world RE applications.

The increasing adoption of Large Language Models (LLMs) has significantly lowered technical barriers across various domains, enabling tasks such as code generation, database querying, and service composition. This work explores the potential of LLMs to facilitate human access to heterogeneous data in Cyber-Physical Systems (CPS). The proposed system interprets natural language queries and, by leveraging in-context learning, dynamically synthesizes data integration pipelines that compose data services, thus enabling access to information from multiple sources. The execution of a pipeline generates a table that fulfills the input query, effectively integrating and structuring the retrieved data. This paper presents the design and implementation of the proposed solution and its evaluation using a real-world case study and the BIRD benchmark dataset. The results demonstrate its effectiveness in improving data retrieval, data service design, and execution in CPS.

This study introduces a novel approach to mine risk factors for short-term death after liver transplantation (LT). The method outputs intelligible survival models by combining Cox’s regression with a genetic programming technique known as multi-objective symbolic regression (MOSR). We consider 485 Electronic Health Records (EHRs) of patients who underwent LT, containing information on hospitalization and preoperative conditions, with a focus on infections and colonizations by multi-resistant Gram-negative bacteria. We evaluate MOSR outcomes against several performance metrics and demonstrate that they are well-calibrated, predictive, safe, and parsimonious. Finally, we select the most promising post-LT early survival risk score based on information criteria, performance, and out-of-distribution safety. Validating this technique at a multicenter level could improve service pipeline logistics through a trustworthy machine-learning method.

Scores are mathematical combinations of elementary indicators (EIs) widely used to measure complex phenomena. Upon the theoretical framework definition, score construction requires a method to aggregate EIs. Aggregation is usually chosen among known methodologies fixing its shape through a try and error approach. Only then are the predictive power, the distribution of the index, and its ability to stratify the population measured. In this paper, we propose a novel data-driven approach that generates analytic aggregation methods relying on multi-objective symbolic regression. We translate the properties that the index must exhibit into optimization goals so that optimal index candidates replicate target variables, data balancing, and stratification. We run experiments on real data sets to solve three main score management problems: data-driven score simplification, generation, and combination. The results obtained show the effectiveness and robustness of the proposed approach.

COVID-19-related pneumonia requires different modalities of Intensive Care Unit (ICU) interventions at different times to facilitate breathing, depending on severity progression. The ability for clinical staff to predict how patients admitted to hospital will require more or less ICU treatment on a daily basis is critical to ICU management. For real datasets that are sparse and incomplete and where the most important state transitions (dismissal, death) are rare, a standard Hidden Markov Model (HMM) approach is insufficient, as it is prone to overfitting. In this paper we propose a more sophisticated ensemble–based approach that involves training multiple HMMs, each specialized in a subset of the state transitions, and then selecting the more plausible predictions either by selecting or combining the models. We have validated the approach on a live dataset of about 1,000 patients from a partner hospital. Our results show that rare events, as well as the transitions to the most severe treatments outperform state of the art approaches.

The digital transformation of organizations is boosting workplace networking and collaboration while making it “observable” with unprecedented timeliness and detail. However, the informational and managerial potential of work datafication is still largely unutilized in Human Resource Management (HRM) and its social benefits, both at the individual and the organizational level, remain largely unexplored. Our research focuses on the relationship between digitally tracked work behaviors and employee attitudes and, in so doing, it explores work datafication as a source of social good. As part of a wider research program, this paper presents some data analysis we performed on a collection of Enterprise Collaboration Software (ECS) data, in search for promising correlations between behavioral and relational (digital) work patterns and employee attitudes. To this end, we transformed the digital actions performed by 106 employees during a one year period into a graph representation to analyze data under two different points of view: the individual (behavioral) perspective, according to the user who performed the action and the action undertaken, and the social (relational) perspective, making explicit the interactions between users and the objects of their actions. Different employees' rankings are thus derived and correlated with their attitudes. We discuss the obtained results and their benefits in terms of perspective social good for both the company and the employee.

The study of such a complex phenomenon as cancer, which depends on several but unexplored and unclear factors, needs new ways to visualize, analyze and combine different data both on species characteristics and genes function. To this respect, we propose a novel platform, named VarCopy, supporting visual Exploratory Data Analysis (EDA) in the context of Copy Number Variation (CNV) data. The platform will be publicly available as a web application soon, and is, to our best knowledge, the first tool allowing visual, interactive exploration and analysis of the CNV landscape of multiple species, allowing the identification of new target genes that might be useful for biomedical research.

The digital transformation of organizations is making workplace collaboration more and more powerful and work always "observable"; however, the informational and managerial potential of the generated data is still largely unutilized in Human Resource Management (HRM). Our research, conducted in collaboration with business engineers and economists, aims at exploring the relationship between digital work behaviors and employee attitudes. This paper is a work-in-progress contribution that presents a preliminary phase of data analysis we performed on a collection of Enterprise Collaboration Software (ECS) data. In the exploratory data analysis step, we analyze data in their original table format and elaborate it according to the user who performed the action and the performed action. Then, we move to a graph representation in order to make explicit the interaction between users and the objects of their actions. Finally, we introduce the concept of employee-attitude-oriented pattern as a mean to derive significant views over the overall graph and discuss Social Network Analysis (SNA) approaches that can be exploited for our purposes.

The technological foundation of smart manufacturing consists of cyber-physical systems and the Internet-of-Things (IoT). Each IoT device in a smart factory can be coupled with a digital twin, that is, a dynamic virtual representation of the physical system across its life-cycle using real-time sensor data. Currently, the manufacturing process itself, the involved devices, and how they interact, is designed by human experts in a traditional way. We envision an architecture where humans can instead specify a goal and take advantage of technologies such as digital twins to automatically compose the corresponding physical processes, sharing some analogies with the notion of Web service composition.

The recently introduced streaming table concept, a fully native representation of streaming data inside a DBMS, enabled modern data-intensive applications with one-time queries (OTQs) and continuous queries (CQs) capabilities on both streaming and standard relational tables. In this paper, we fully acknowledge the temporal nature of streaming tables and we propose to go one step further and integrate them in a temporal DBMS context, where time management is native. Our aim is to break the traditional barrier between the streaming and the temporal worlds, offering complete interoperability between streams and temporal data. To this end, we present a continuous temporal algebra supporting both OTQs and CQs seamlessly on streaming, standard and temporal relational tables. We further show how the transition from continuous to one-time semantics can be managed by defining suitable translation rules, which can also be used as a basis for the implementation of the proposed continuous algebra in a temporal DBMS.

When dealing with a specific patient case, physicians are often interested in retrieving a personalized version of a clinical guideline, that is a version tailored to their use needs. In a patient-centric scenario, empowered patients make up another class of users interested in retrieving personalized care plans from a guideline repository. In their previous work, the authors proposed techniques to efficiently provide ontology-based personalized access to very large collections of multi-version clinical guidelines. In this paper, they address the problem of also dealing with a multi-version ontology used to support personalized access to clinical guidelines. The authors' approach allows the semantic indexing of guideline contents with respect to multi-version ontology classes and exploits the IS-A relationship among such classes for granting personalized access. Efficiency is ensured by a newly introduced annotation scheme for guidelines and solutions to cope with the evolution of ontology structure. The tests performed on a prototype implementation confirm the goodness of the approach.

Data stream management systems (DSMSs) are conceived for running continuous queries (CQs) on the most recently streamed data. This model does not completely fit the needs of several modern data-intensive applications that require to manage recent/historical/static data and execute both CQs and OTQs joining such data. In order to cope with these new needs, some DSMSs have moved toward the integration of database management systems (DBMSs) functionalities to augment their capabilities. In this paper we adopt the opposite perspective and we lay the groundwork for extending DBMSs to natively support streaming facilities. To this end, we introduce a new kind of table, the streaming table, as a persistent structure where streaming data enters and remains stored for a long period, ideally forever. Streaming tables feature a novel access paradigm: continuous writes and one-time as well as continuous reads. We present a streaming table implementation and two novel types of indices that efficiently support both update and scan high rates. A detailed experimental evaluation shows the effectiveness of the proposed technology.

In this paper we investigate the employment of different data reduction techniques to minimize V2I communication in an Intelligent Transportation System (ITS). We consider the context of the PEGASUS Project, where vehicles are equipped with sensor-based devices able to compute and communicate to a Control Centre (CC) information like vehicleăĂŹs position and speed. The CC relies on a general-purpose data management module that supports the execution of continuous queries as well as standard SQL one-time queries on the collected data to provide various infomobility services.

Inter-business collaborative contexts prefigure a distributed scenario where companies organize and coordinate themselves to develop common and shared opportunities. Traditional business intelligence systems do not provide support to this end. Peer Data Management Systems (PDMSs) have been proposed as architectures to support sharing of operational data across networks of peers while guaranteeing peers' autonomy, based on semantic mappings that mediate between the heterogeneous schemata exposed by peers. In line with the PDMS infrastructure, in this paper we envision a peer-to-peer data warehousing architecture based on a network of heterogeneous peers, each exposing query answering functionalities aimed at sharing business information. To enhance the decision making process, an OLAP query expressed on a peer needs be properly reformulated on the other peers. In this direction, we present a language for the definition of mappings between the multidimensional schemata of peers, and we introduce a query reformulation framework that relies on the translation of these mappings towards relational schemata. Finally, we sketch the query reformulation algorithm by outlining the reformulation steps of typical OLAP queries.

The largeness and the heterogeneity of most graph-modeled datasets in several database application areas make the query process a real challenge because of the lack of a complete knowledge of the vocabulary used, as well as of the information about the structural relationships between the data.

Peer Data Management Systems (PDMSs) have been introduced as a solution to the problem of large-scale sharing of semantically rich data. A PDMS consists of semantic peers connected through semantic mappings. Querying a PDMS may lead to very poor results, because of the semantic degradation due to the approximations given by the traversal of the semantic mappings, thus leading to the problem of how to boost a network of mappings in a PDMS.In this paper we propose a strategy for the incremental maintenance of a flexible network organization that clusters together peers which are semantically related in Semantic Overlay Networks (SONs), while maintaining a high degree of node autonomy. Semantic features, a summarized representation of clusters, are stored in a light structure which effectively assists a newly entering peer when choosing its semantically closest overlay networks. Then, each peer is supported in the selection of its own neighbors within each overlay network according to two policies: Range-based selection and k-NN selection. For both policies, we introduce specific algorithms which exploit a distributed indexing mechanism for efficient network navigation. The proposed approach has been implemented in a prototype where its effectiveness and efficiency have been extensively tested.

In this paper, we propose a versatile disambiguation approach which can be used to make explicit the meaning of structure based information such as XML schemas, XML document structures, web directories, and ontologies. It can be of support to the semantic-awareness of a wide range of applications, from schema matching and query rewriting to peer data management systems, from XML data clustering to ontology-based automatic annotation of web pages and query expansion. The effectiveness of the achieved results has been experimentally proved and is founded both on a flexible exploitation of the structure context, whose extraction can be tailored on the specific application needs, and of the information provided by commonly available thesauri such as WordNet.

Textual data is the main electronic form of knowledge representation. Sentences, meant as logic units of meaningful word sequences, can be considered its backbone. In this paper, we propose a solution based on a purely syntactic approach for searching similarities within sentences, named approximate sub2sequence matching. This process being very time consuming, efficiency in retrieving the most similar parts available in large repositories of textual data is ensured by making use of new filtering techniques. As far as the design of the system is concerned, we chose a solution that allows us to deploy approximate sub2 sequence matching without changing the underlying database.

Management of indeterminate temporal expressions is useful in a wide range of applications, from designing and querying temporal databases to knowledge representation and reasoning in artificial intelligence. In this paper, we focus on the representation and management of indeterminate dates, corresponding to a common use of temporal indeterminacy which can be found in (historical) texts written in natural language, as in expressions like: around 1624, near the end of the fourteenth century, etc. In this context, we adapt and improve the probabilistic approach designed for the TSQL2 language and further developed by Dyreson and Snodgrass, and show how it can be effectively and efficiently adopted for the management of indeterminate dates.