Key information extraction (KIE) from visually rich documents, such as receipts and forms, involves a deep understanding of textual, visual, and layout feature information. Transformers fine-tuned for KIE achieve state-of-the-art performance but lack generality and portability across different domains. In contrast, vision large language models (VLLMs) offer higher flexibility and zero-shot capability but fall short with domain-specific layout relations unless performing a resource-demanding supervised fine-tuning. To reach the best compromise solution between lightweight models and VLLMs, we propose KIEPrompter, a cost-effective LLM-based KIE approach that leverages the predictions of lightweight models as external knowledge injected into VLLM prompts. By incorporating these auxiliary predictions, VLLMs are guided to attend relevant multimodal content without ad hoc training. The accuracy results achieved by KIEPrompter in three benchmark document collections are superior to those of VLLMs in both zero-shot and layout-sensitive scenarios. We compare various strategies for incorporating lightweight model predictions, ranging from coarse-grained predictions without explicit confidence scores to fine-grained per-element network logits. We also demonstrate that our approach is robust to the absence of specific classes in trained lightweight models, as the VLLMs' pre-training compensates for the limited generality of lightweight models.

Forecasting surface water dynamics is crucial for water resource management and climate change adaptation. However, the field lacks comprehensive datasets and standardized benchmarks. In this paper, we introduce HydroChronos, a large-scale, multi-modal spatiotemporal dataset for surface water dynamics forecasting designed to address this gap. We couple the dataset with three forecasting tasks. The dataset includes over three decades of aligned Landsat 5 and Sentinel-2 imagery, climate data, and Digital Elevation Models for diverse lakes and rivers across Europe, North America, and South America. We also propose AquaClimaTempo UNet, a novel spatiotemporal architecture with a dedicated climate data branch, as a strong benchmark baseline. Our model significantly outperforms a Persistence baseline for forecasting future water dynamics by +14% and +11% F1 across change detection and direction of change classification tasks, and by +0.1 MAE on the magnitude of change regression. Finally, we conduct an Explainable AI analysis to identify the key climate variables and input channels that influence surface water change, providing insights to inform and guide future modeling efforts.

Density-based clustering focuses on defining clusters consisting of contiguous regions characterized by similar densities of points. Traditional approaches identify core points first, whereas more recent ones initially identify the cluster borders and then propagate cluster labels within the delimited regions. Both strategies encounter issues in presence of multi-density regions or when clusters are characterized by noisy borders. To overcome the above issues, we present a new clustering algorithm that relies on the concept of bridge point. A bridge point is a point whose neighborhood includes points of different clusters. The key idea is to use bridge points, rather than border points, to partition points into clusters. We have proved that a correct bridge point identification yields a cluster separation consistent with the expectation. To correctly identify bridge points in absence of a priori cluster information we leverage an established unsupervised outlier detection algorithm. Specifically, we empirically show that, in most cases, the detected outliers are actually a superset of the bridge point set. Therefore, to define clusters we spread cluster labels like a wildfire until an outlier, acting as a candidate bridge point, is reached. The proposed algorithm performs statistically better than state-of-the-art methods on a large set of benchmark datasets and is particularly robust to the presence of intra-cluster multiple densities and noisy borders.

The ability to correctly identify areas damaged by forest wildfires is essential to plan and monitor the restoration process and estimate the environmental damages after such catastrophic events. The wide availability of satellite data, combined with the recent development of machine learning and deep learning methodologies applied to the computer vision field, makes it extremely interesting to apply the aforementioned techniques to the field of automatic burned area detection. One of the main issues in such a context is the limited amount of labeled data, especially in the context of semantic segmentation. In this paper, we introduce a publicly available dataset for the burned area detection problem for semantic segmentation. The dataset contains 73 satellite images of different forests damaged by wildfires across Europe with a resolution of up to 10m per pixel. Data were collected from the Sentinel-2 L2A satellite mission and the target labels were generated from the Copernicus Emergency Management Service (EMS) annotations, with five different severity levels, ranging from undamaged to completely destroyed. Finally, we report the benchmark values obtained by applying a Convolutional Neural Network on the proposed dataset to address the burned area identification problem.

Discovering patterns that represent key spatial or temporal dependencies among data is a well-known exploratory data mining task. However, prior works either separately analyze spatial and temporal dependencies or discover joint spatiotemporal properties of specific trajectories observed over a region of interest. With the goal of generalizing the information provided by spatiotemporal patterns, in this paper we extract sequences of discrete events showing spatiotemporally invariant properties. We seek patterns whose corresponding instances in the source data differ only due to an invariant spatiotemporal transformation. We denote such a new type of patterns as SpatioTemporally Invariant. We also propose an efficient algorithm to mine STInvs and validate its efficiency and effectiveness on real data.

Recent technological advancements have enabled generating and collecting huge amounts of data in a daily manner. This data is used for different purposes that may impact us on an unprecedented scale. Understanding the data, including detecting its outliers, is a critical step before utilizing it.Outlier detection has been studied well in the literature but the existing approaches fail to scale to these very large settings. In this paper, we propose DBSCOUT, an efficient exact algorithm for outlier detection with a linear complexity that can run in parallel over multiple independent machines, making it a fit for the settings with billions of tuples. Besides the theoretical analysis, our experiment results confirm orders of magnitude improvement over the existing work, proving the efficiency, scalability, and effectiveness of our approach.

Uncontrolled wildfires are dangerous events capable of harming people safety. To contrast their increasing impact in recent years, a key task is an accurate detection of the affected areas and their damage assessment from satellite images. Current state-of-the-art solutions address such problem through a double convolutional neural network able to automatically detect wildfires in satellite acquisitions and associate a damage index from a defined scale. However, such deep-learning model performance is strongly dependent on many factors. In this work, we specifically focus on a key parameter, i.e., the loss function, exploited in the underlying neural networks. Besides the state-of-the-art solutions based on the Dice-MSE, among the many loss functions proposed in literature, we focus on the Binary Cross-Entropy (BCE) and the Intersection over Union (IoU), as two representatives of the distribution-based and region-based categories, respectively. Experiments show that the BCE loss function coupled with a double-step U-Net architecture provides better results than current state-of-the-art solutions on a public labeled dataset of European wildfires.

Sentence-based summarization aims at extracting concise summaries of collections of textual documents. Summaries consist of a worthwhile subset of document sentences. The most effective multilingual strategies rely on Latent Semantic Analysis (LSA) and on frequent itemset mining, respectively. LSA-based summarizers pick the document sentences that cover the most important concepts. Concepts are modeled as combinations of single-document terms and are derived from a term-by-sentence matrix by exploiting Singular Value Decomposition (SVD). Itemset-based summarizers pick the sentences that contain the largest number of frequent itemsets, which represent combinations of frequently co-occurring terms. The main drawbacks of existing approaches are (i) the inability of LSA to consider the correlation between combinations of multiple-document terms and the underlying concepts, (ii) the inherent redundancy of frequent itemsets because similar itemsets may be related to the same concept, and (iii) the inability of itemset-based summarizers to correlate itemsets with the underlying document concepts. To overcome the issues of both of the abovementioned algorithms, we propose a new summarization approach that exploits frequent itemsets to describe all of the latent concepts covered by the documents under analysis and LSA to reduce the potentially redundant set of itemsets to a compact set of uncorrelated concepts. The summarizer selects the sentences that cover the latent concepts with minimal redundancy. We tested the summarization algorithm on both multilingual and English-language benchmark document collections. The proposed approach performed significantly better than both itemset- and LSA-based summarizers, and better than most of the other state-of-the-art approaches.

Clustering big data is a challenging task, because the majority of high-quality clustering algorithms do not scale well with respect to the data set cardinality. To tackle the scalability problem, we propose a general-purpose density-based preprocessing technique, called SCOUT, implemented in the Spark framework. It allows compacting the original data by means of a set of representative points, while still preserving the original data distribution and density information. This small set of representative points may become the input to almost any clustering algorithm. Thus, also complex, high-quality in-memory algorithms can be applied. A thorough experimental evaluation shows that the proposed approach is efficient and at the same time effective.

In the last years, we are witnessing an increasing availability of geolocated data, ranging from satellite images to user generated content (e.g., tweets). This big amount of data is exploited by several cloud-based applications to deliver effective and customized services to end users. In order to provide a good user experience, a low-latency response time is needed, both when data are retrieved and provided. To achieve this goal, current geospatial applications need to exploit efficient and scalable geospatial databases, the choice of which has a high impact on the overall performance of the deployed applications. In this paper, we compare, from a qualitative point of view, four state-of-the-art SQL and NoSQL databases with geospatial features, and then we analyze the performances of two of them, selecting the ones based on the Database-as-a-service (DBaaS) model: Azure SQL Database and Azure DocumentDB (i.e., an SQL database versus a NoSQL one). The empirical evaluation shows pros and cons of both solutions and it is performed on a real use case related to an emergency management application.

Physicians and health care organizations always collect large amounts of data during patient care. These large and high-dimensional datasets are usually characterized by an inherent sparseness. Hence, analyzing these datasets to figure out interesting and hidden knowledge is a challenging task. This article proposes a new data mining framework based on generalized association rules to discover multiple-level correlations among patient data. Specifically, correlations among prescribed examinations, drugs, and patient profiles are discovered and analyzed at different abstraction levels. The rule extraction process is driven by a taxonomy to generalize examinations and drugs into their corresponding categories. To ease the manual inspection of the result, a worthwhile subset of rules (i.e., nonredundant generalized rules) is considered. Furthermore, rules are classified according to the involved data features (medical treatments or patient profiles) and then explored in a top-down fashion: from the small subset of high-level rules, a drill-down is performed to target more specific rules. The experiments, performed on a real diabetic patient dataset, demonstrate the effectiveness of the proposed approach in discovering interesting rule groups at different abstraction levels.

Multidocument summarization addresses the selection of a compact subset of highly informative sentences, i.e., the summary, from a collection of textual documents. To perform sentence selection, two parallel strategies have been proposed: (a) apply general-purpose techniques relying on data mining or information retrieval techniques, and/or (b) perform advanced linguistic analysis relying on semantics-based models (e.g., ontologies) to capture the actual sentence meaning. Since there is an increasing need for processing documents written in different languages, the attention of the research community has recently focused on summarizers based on strategy (a). This article presents a novel multilingual summarizer, namely MWI-Sum (Multilingual Weighted Itemset-based Summarizer), that exploits an itemset-based model to summarize collections of documents ranging over the same topic. Unlike previous approaches, it extracts frequent weighted itemsets tailored to the analyzed collection and uses them to drive the sentence selection process. Weighted itemsets represent correlations among multiple highly relevant terms that are neglected by previous approaches. The proposed approach makes minimal use of language-dependent analyses. Thus, it is easily applicable to document collections written in different languages. Experiments performed on benchmark and real-life collections, English-written and not, demonstrate that the proposed approach performs better than state-of-the-art multilingual document summarizers.

Frequent weighted itemsets represent correlations frequently holding in data in which items may weight differently. However, in some contexts, e.g., when the need is to minimize a certain cost function, discovering rare data correlations is more interesting than mining frequent ones. This paper tackles the issue of discovering rare and weighted itemsets, i.e., the infrequent weighted itemset (IWI) mining problem. Two novel quality measures are proposed to drive the IWI mining process. Furthermore, two algorithms that perform IWI and Minimal IWI mining efficiently, driven by the proposed measures, are presented. Experimental results show efficiency and effectiveness of the proposed approach.

A promising approach to Bayesian classification is based on exploiting frequent patterns, i.e., patterns that frequently occur in the training data set, to estimate the Bayesian probability. Pattern-based Bayesian classification focuses on building and evaluating reliable probability approximations by exploiting a subset of frequent patterns tailored to a given test case. This paper proposes a novel and effective approach to estimate the Bayesian probability. Differently from previous approaches, the Entropy-based Bayesian classifier, namely EnBay, focuses on selecting the minimal set of long and not overlapped patterns that best complies with a conditional-independence model, based on an entropy-based evaluator. Furthermore, the probability approximation is separately tailored to each class. An extensive experimental evaluation, performed on both real and synthetic data sets, shows that EnBay is significantly more accurate than most state-of-the-art classifiers, Bayesian and not.

Associative classification is characterized by accurate models and high model generation time. Most time is spent in extracting and postprocessing a large set of irrelevant rules, which are eventually pruned. We propose I-prune, an item-pruning approach that selects uninteresting items by means of an interestingness measure and prunes them as soon as they are detected. Thus, the number of extracted rules is reduced and model generation time decreases correspondingly. A wide set of experiments on real and synthetic data sets has been performed to evaluate I-prune and select the appropriate interestingness measure. The experimental results show that I-prune allows a significant reduction in model generation time, while increasing (or at worst preserving) model accuracy. Experimental evaluation also points to the chi-square measure as the most effective interestingness measure for item pruning. © 2012 Wiley Periodicals, Inc.

Singular Value Decomposition (SVD) has been extensively used in the classification context as a preprocessing step aiming to reduce the number of features of the input space. Traditional classification algorithms are then applied on the new space to generate accurate models. In this paper, we propose a different use of SVD. In our approach SVD is the building block of a new classification algorithm, called CMF, and not that of a feature reduction algorithm. In particular, we propose a new classification algorithm where the classification model corresponds to the k largest right singular vectors of the factorization of the training dataset obtained by applying SVD. The selected singular vectors allows representing the main "characteristics" of the training data and can be used to provide accurate predictions. The experiments performed on 15 structured UCI datasets show that CMF is efficient and, despite its simplicity, it is more accurate than many state of the art classification algorithms.

This paper presents a novel semi-automatic approach to construct conceptual ontologies over structured data by exploiting both the schema and content of the input dataset. It effectively combines two well-founded database and data mining techniques, i.e., functional dependency discovery and association rule mining, to support domain experts in the construction of meaningful ontologies, tailored to the analyzed data, by using Description Logic (DL). To this aim, functional dependencies are first discovered to highlight valuable conceptual relationships among attributes of the data schema (i.e., among concepts). The set of discovered correlations effectively support analysts in the assertion of the Tbox ontological statements (i.e., the statements involving shared data conceptualizations and their relationships). Then, the analyst-validated dependencies are exploited to drive the association rule mining process. Association rules represent relevant and hidden correlations among data content and they are used to provide valuable knowledge at the instance level. The pushing of functional dependency constraints into the rule mining process allows analysts to look into and exploit only the most significant data item recurrences in the assertion of the Abox ontological statements (i.e., the statements involving concept instances and their relationships).

Associative classification is a promising technique to build accurate classifiers. However, in large or correlated datasets, association rule mining may yield huge rule sets. Hence, several pruning techniques have been proposed to select a small subset of high quality rules. We argue that rule pruning should be reduced to a minimum, since the availability of a "rich" rule set may improve the accuracy of the classifier. The L^3 associative classifier is built by means of a lazy pruning technique which discards exclusively rules that only misclassify training data. Classification of unlabeled data is performed in two steps. A small subset of high quality rules is first considered. When this set is not able to classify the data, a larger rule set is exploited. This second set includes rules usually discarded by previous approaches. To cope with the need of mining large rule sets and efficiently use them for classification, a compact form is proposed to represent a complete rule set in a space-efficient way and without information loss. An extensive experimental evaluation on real and synthetic datasets shows that L^3 improves the classification accuracy with respect to previous approaches.

XML is a rather verbose representation of semistructured data, which may require huge amounts of storage space. We propose a summarized representation of XML data, based on the concept of instance pattern, which can both provide succinct information and be directly queried. The physical representation of instance patterns exploits itemsets or association rules to summarize the content of XML datasets. Instance patterns may be used for (possibly partially) answering queries, either when fast and approximate answers are required, or when the actual dataset is not available, for example, it is currently unreachable. Experiments on large XML documents show that instance patterns allow a significant reduction in storage space, while preserving almost entirely the completeness of the query result. Furthermore, they provide fast query answers and show good scalability on the size of the dataset, thus overcoming the document size limitation of most current XQuery engines.

Associative classification is a promising technique for the generation of highly precise classifiers. Previous works propose several clever techniques to prune the huge set of generated rules, with the twofold aim of selecting a small set of high quality rules, and reducing the chance of overfitting. In this paper, we argue that pruning should be reduced to a minimum and that the availability of a large rule base may improve the precision of the classifier without affecting its performance. In L/sup 3/ (Live and Let Live), a new algorithm for associative classification, a lazy pruning technique iteratively discards all rules that only yield wrong case classifications. Classification is performed in two steps. Initially, rules which have already correctly classified at least one training case, sorted by confidence, are considered If the case is still unclassified, the remaining rules (unused during the training phase) are considered, again sorted by confidence. Extensive experiments on 26 databases from the UCI machine learning database repository show that L/sup 3/ improves the classification precision with respect to previous approaches.