Heterogeneous Sources Research Articles

Protecting privacy in the age of big data: exploring data linking methods for quasi-identifier selection

This paper explores the limitations faced by current solutions for selecting quasi-identifying attributes in the context of Privacy-Preserving Data Publishing (PPDP). PPDP stipulates that any published personal data should not be linkable to other available data sources in a manner that could potentially lead to individual re-identification or compromise sensitive data. The state-of-the-art methods for selecting quasi-identifying attributes commonly rely on heuristic evaluations to assess the risk of re-identification associated with each attribute. We hypothesize that these heuristic-based methods could be significantly improved by complementing them with empirical methods capable of quantifying the external linkability of dataset attributes. This empirical layer would enable a fine-tuning of the obfuscation of attributes within the dataset, thereby preventing the unnecessary privatization of attributes beyond potential attackers’ reach while ensuring privatization of those easily accessible. For this purpose, we explore recent advancements in identifying semantically related datasets across heterogeneous data sources. Although initially developed for purposes beyond privacy preservation, these methods support our initiative by uncovering potential links with external data and thus providing empirical evidence for the identification of attributes as quasi-identifiers. Finally, we discuss potential pathways to implement this empirical layer in quasi-identifier identification systems.

Interoperable information modelling leveraging asset administration shell and large language model for quality control toward zero defect manufacturing

In the era of Industry 4.0, Zero Defect Manufacturing (ZDM) has emerged as a prominent strategy for quality improvement, emphasizing data-driven approaches for defect prediction, prevention, and mitigation. The success of ZDM heavily depends on the availability and quality of data typically collected from diverse and heterogeneous sources during production and quality control, presenting challenges in data interoperability. Addressing this, we introduce a novel approach leveraging Asset Administration Shell (AAS) and Large Language Models (LLMs) for creating interoperable information models that incorporate semantic contextual information to enhance the interoperability of data integration in the quality control process. AAS, initiated by German industry stakeholders, shows a significant advancement in information modeling, blending ontology and digital twin concepts for the virtual representation of assets. In this work, we develop a systematic, use-case-driven methodology for AAS-based information modeling. This methodology guides the design and implementation of AAS models, ensuring model properties are presented in a unified structure and reference external standardized vocabularies to maintain consistency across different systems. To automate this referencing process, we propose a novel LLM-based algorithm to semantically search model properties within a standardized vocabulary repository. This algorithm significantly reduces manual intervention in model development. A case study in the injection molding domain demonstrates the practical application of our approach, showcasing the integration and linking of product quality and machine process data with the help of the developed AAS models. Statistical evaluation of our LLM-based semantic search algorithm confirms its efficacy in enhancing data interoperability. This methodology offers a scalable and adaptable solution for various industrial use cases, promoting widespread data interoperability in the context of Industry 4.0.

Heterogeneous Sources, Distribution, and removal processes of dissolved black carbon from East China Sea shelf to open ocean of Northwest Pacific

Absract: Dissolved black carbon (DBC) can be transported from land to the ocean, representing a significant slow-cycling component in the global carbon cycle. However, the source, distribution, and degradation of DBC in the ocean are not well understood. In this study, water samples were collected from the East China Sea (ECS) shelf to the Northwest Pacific Ocean. The benzene polycarboxylic acid (BPCA) method and ultraviolet–visible spectroscopy were used to analyze DBC and dissolved organic matter, respectively. The concentration of DBC and the (B6CA + B5CA)/BPCAs ratio in surface water showed a decreasing trend from the ECS inner shelf towards the open ocean. This trend indicates the preferential removal of condensed aromatic compounds by photodegradation, while also implicating dilution by open ocean water as a significant factor contributing to the reduction of DBC concentrations. A significant negative correlation between DBC concentration and salinity was observed in surface water samples (r = − 0.88; p < 0.01). However, this relationship weakened and disappeared in offshore samples, suggesting a reduced influence of riverine inputs. Furthermore, vertical profiles of water samples (0 – 1000 m) revealed a decrease in DBC concentrations and an increase in the ratio of (B6CA + B5CA)/BPCAs, indicating the preferential photodegradation of condensed aromatic DBC molecules in the euphotic zone and the selective accumulation of these molecules in the ocean interior. The DBC flux into the Northwest Pacific Ocean was estimated to be 1.46 Tg/year with 51.7 to 73.1 % removed through photodegradation, while the remaining DBC was either transported into the deep ocean or biodegraded. Given the important role of photodegradation in regulating the DBC flux in the Northwest Pacific Ocean, we recommend further research into its impact on the refractory carbon cycle in global oceans.

Triassic–Jurassic alkaline magmatism in the Ivrea-Verbano Zone, Southern Alps: A zircon perspective on mantle sources and geodynamic significance

The Ivrea-Verbano Zone (IVZ; western Southern Alps) consists of a distinctive sequence of the lower continental crust of the Adriatic plate, extending down to the subcontinental lithospheric mantle. It is characterized by a large variety of intrusive bodies of variable geochemical composition and age, offering a unique insight into the evolution of mantle-derived magmatism in post-collisional and extensional geodynamic settings. In this study, we characterize a suite of alkaline dykes intruding a mantle massif – the Finero Phlogopite Peridotite in northern IVZ. These dykes include zircon-bearing diorites and anorthosites, mainly composed by HFSE-rich amphibole and phlogopite, albite (>90 vol% in anorthosites) and apatite. Zircon, monazite, ilmenite, titanite, Nb-rich oxides, and carbonates are common accessory minerals. Additionally, a “composite” diorite dyke containing both HFSE-rich and HFSE-poor amphiboles was investigated. The study is aimed at providing new trace element, UPb geochronological and LuHf isotopes dataset on zircon from these alkaline dykes, to refine the understanding of their mantle source characteristics, emplacement age and geodynamic implications. The trace element composition of zircons from the studied dykes points to segregation from melts with alkaline to ultra-alkaline affinity. Concordia UPb ages of zircon from the alkaline diorite dykes span from 221 to 191 Ma, which are interpreted as the result of multiple crystallization/recrystallization stages related to different magmatic pulses. Conversely, zircon from anorthosite and composite diorite dykes yield a narrow time range of 198–202 Ma, highlighting the occurrence of a magmatic pulse around ca. 200 Ma. The εHf(t) values (+13.4 to +5.7) of zircon from the alkaline diorite dykes are significantly more positive compared to the values in those from anorthosite and composite diorite dykes (+4.2 to −0.4), suggesting that the parental melts were derived from heterogeneous asthenospheric mantle sources with low to moderate amount of recycled continental crust components. Our data and reappraisal of the literature indicate that the IVZ experienced a protracted period of alkaline magmatism from ca. 235 Ma to 185 Ma. The melts migrated along mantle shear zones during the Triassic-Jurassic lithospheric extension. Different pulses of alkaline magmatism were associated with relevant extensional tectonic stages recorded by the continental crust, presumably triggering a passive uplift of heterogenous asthenospheric reservoirs over a front of at least 500 km. This tectono-magmatic cycle is a precursor to the focused rifting stage which caused the opening of the Jurassic Alpine Tethys, enhancing Pangea fragmentation.

Joint Short-term Power Forecasting of Hydro-Wind-Photovoltaic Considering Spatiotemporal Delay of Weather Processes

Regional integrated energy systems (RIES) represented by hydro, wind, and solar energy are crucial avenues for future clean energy development, fully leveraging their complementarity can facilitate the orderly supply of renewable energy, significantly enhance the level of new energy consumption, and make outstanding contributions to achieving carbon neutrality goals. However, run-of-river hydropower (RHP), wind power (WP), and photovoltaic (PV) are affected by the chaotic characteristics of the weather system, with significant random fluctuations, and their output characteristics have significant heterogeneity, which brings a big challenge for the complementary coordinated scheduling. The high-accuracy power prediction of RHP, WP, and PV in the region is one of the key means for solving the above problems. To meet this challenge, by analyzing the spatio-temporal correlation properties between weather processes and station output, a novel joint prediction framework for short-term power forecasting (STPF) of regional hydro-wind-PV clusters is proposed. Firstly, to solve the problem of significant heterogeneity of water, wind, and solar, a Differentiated Spatio-Temporal Mixture Network (DSTMN) is proposed, which differentially encodes the weather and power information at multiple locations, enabling the extraction of common features while preserving their specificities, which addresses the difficulty of representing the inherent correlation patterns between wide-area meteorological information and heterogeneous energy sources. Secondly, to further explore the spatio-temporal correlation between meteorological information and power information, to improve the forecasting performance of the model, a spatio-temporal-lagged correlation (STLC) that can quantify their spatio-temporal delays is proposed. By analyzing the correlation characteristics between regional grid numerical weather prediction (NWP) information and station output at different spatial and temporal scales, the optimal spatial location and delay time of NWP inputs under the current scenario are determined for each station. Based on this, the best NWP scene probabilistic transfer model is established based on the Rank Bayesian Ensemble (RBE) method. By fitting the conditional probability distribution of the current best NWP spatiotemporal location and the 2nd day’s best NWP spatiotemporal location, providing a reliable input for the STPF model. Finally, a case study with 164 hydro-wind-solar stations in China demonstrates the effectiveness of the proposed method. Specifically, we achieved an accuracy improvement of 0.46% — 11.03% on the 2nd day of forecasting compared to benchmark methods.

An IoT-based data analysis system: A case study on tomato cultivation under different irrigation regimes

The exploitation of modern technologies in heterogeneous farming scenarios with different crops cultivation is nowadays an effective solution to implement the concept of Smart Agriculture (SA). Following this approach, in this study the tomato plants’ response to different irrigation regimes is investigated through the implementation of an Internet of Things (IoT)-oriented SA data collection and monitoring system. In particular, the experimentation is conducted on tomatoes grown at three different irrigation regimes: namely, at 100%, 60%, and 30% of the Italian irrigation recommendation service, denoted as Irriframe. The proposed platform, denoted as Agriware, is able to: (i) evaluate information from heterogeneous data sources, (ii) calculate agronomic indicators (e.g., Growing Degree Days, GDD), and (iii) monitor on-field parameters (e.g., water consumption). Different plant-related parameters have been collected to assess the response to water stress (e.g., Soil Plant Analysis Development (SPAD), chlorophyll content, fluorescence, and others), along with leaf color and final production evaluations. The obtained results show that the best irrigation regime, in terms of plant health and productivity, corresponds to 60% of Irriframe, allowing significant water savings for the cultivation.

Paleoproterozoic sediment-derived magmas reveal a late orogenic stage in the southern São Francisco Craton, Brazil: Evidence from petrogenesis of the 2.0 Ga Cupim Pluton leucogranites

Sediment-derived granites (S-type) can provide detailed information about continental crust reworking processes and tectonic regimes through time. Their compositions reflect the nature of sedimentary rocks that melted to form them. Inherited accessory minerals in these granites can provide information regarding the age and nature of the crustal rocks that contributed to the formation of their sedimentary source rock. In addition, S-type granites typically form during collisional stages in modern orogens, and their Precambrian records can provide clues about ancient tectonics. In this contribution, we present a detailed petrogenesis of the Cupim Pluton, located in the Paleoproterozoic Mineiro Belt (Brazil). The Cupim pluton is composed of two-mica and garnet-bearing leucogranites that are strongly peraluminous (ASI > 1.1), high-K (K2O/Na2O > 1), low MgO + FeO + TiO2 (<2 wt%). Its composition is similar to that of melts experimentally produced by muscovite/biotite dehydration melting reactions. Rb/Sr and Rb/Ba ratios suggest that these leucogranites are formed by the melting of heterogeneous sedimentary source(s). Zircon crystals from the Cupim Pluton unveil a protracted crustal melting period (2.01–1.98 Ga) hitherto unknown in the Mineiro Belt, which itself discloses a late collisional orogenic stage in southern São Francisco Craton. Igneous zircon crystals exhibit εHf(t) between −16.4 and − 0.6, and δ30Si between −1.16 and − 0.26 ‰, which is typical for zircon from sediment-derived granites. Hafnium isotope variations reflect mixing between at least two crustal sources, which is in line with inherited zircon cores that evidence major Archean and minor Paleoproterozoic sediment contributions. The emplacement of the Cupim Pluton leucogranites ends a subduction-collision cycle, following a long period of TTG-sanukitoid magma production, and promoted the stabilization of the southern São Francisco Craton.

Himalayan Leucogranites: A Geochemical Perspective

The geochemical characterization of Himalayan leucogranites offers important insights into both their petrogenesis and Himalayan orogenic processes. Himalayan leucogranites are characterized by strongly peraluminous compositions that are comparable to melts derived from anatexis of sedimentary rocks. Their radiogenic (Sr, Nd, and Hf) isotopic compositions point to metasediments from the Higher Himalayan Sequence of the Indian plate as the primary source rocks, with minor contributions from other lithologies. Himalayan leucogranites display considerable variability in trace element ratios (e.g., Rb/Sr, Nb/Ta, Zr/Hf, and Eu/Eu*) and significant fractionation of non-traditional stable isotopes (e.g., Mg, K, Zn, Rb, and Ba), which provide key constraints on the respective roles of source heterogeneity, crystal fractionation, magma–fluid interaction, and crustal melting in their formation.

Bayesian Inverse Transfer in Evolutionary Multiobjective Optimization

Transfer optimization enables data-efficient optimization of a target task by leveraging experiential priors from related source tasks. This is especially useful in multiobjective optimization settings where a set of tradeoff solutions is sought under tight evaluation budgets. In this article, we introduce a novel concept of inverse transfer in multiobjective optimization. Inverse transfer stands out by employing Bayesian inverse Gaussian process models to map performance vectors in the objective space to population search distributions in task-specific decision space, facilitating knowledge transfer through objective space unification . Building upon this idea, we introduce the first Inverse Transfer Evolutionary Multiobjective Optimizer (invTrEMO). A key highlight of invTrEMO is its ability to harness the common objective functions prevalent in many application areas, even when decision spaces do not precisely align between tasks. This allows invTrEMO to uniquely and effectively utilize information from heterogeneous source tasks as well. Furthermore, invTrEMO yields high-precision inverse models as a significant byproduct, enabling the generation of tailored solutions on-demand based on user preferences. Empirical studies on multi- and many-objective benchmark problems, as well as a practical case study, showcase the faster convergence rate and modeling accuracy of the invTrEMO relative to state-of-the-art evolutionary and Bayesian optimization algorithms. The source code of the invTrEMO is made available at https://github.com/LiuJ-2023/invTrEMO .

Early Cretaceous to Eocene magmatic evolution of Livingston Island, South Shetland Archipelago: Geochronological and isotope geochemical constraints from intrusive suites

The magmatic products exposed across Livingston Island comprise a part of the South Shetland magmatic arc built on an active continental margin. The main magmatic phases of Livingston Island include the Early to Late Cretaceous plutonic and volcanic rock associations, the Eocene plutons, and some sporadically distributed intrusions of Paleocene to Eocene dykes. The intrusions were emplaced into the strongly deformed turbiditic sedimentary rocks of the Miers Bluff Formation. 40Ar39Ar and zircon UPb geochronologic data presented here define four distinct episodes of intrusive activity represented by the emplacement of the Hesperides (125 Ma), the Cerro Mirador (104–97 Ma), the Siddons Point (80 Ma) and the Barnard Point suites (42 Ma), in addition to some dyke intrusions with emplacement ages of 105 to 40 Ma. The intrusions are represented by mafic to intermediate compositions characterized by middle potassic, calc-alkaline affinities, with only a few samples from the Eocene activity exhibiting tholeiitic character. Trace element, Sr-Nd-Pb isotope and zircon Hf isotope variations indicate magma generation in a subduction setting with extensive involvement of continental crust-derived melts for the Early to Late Cretaceous activity. The effects of crustal input on melt chemistry appear to have diminished in time, especially from the Paleocene onwards, leading to magma generation with typical oceanic island arc geochemical signatures. This change in melt chemistry is likely partly related to the roll-back induced overriding plate extension which resulted in a time-integrated increase in the rate of melt movement and a decrease in the residence time in crustal levels. Slab steepening caused by the systematic decrease in plate convergence rate led to an enhanced involvement of isotopically depleted mantle component, producing a heterogeneous source with variable subduction signature.

Kinetic Mg[sbnd]Fe isotope fractionation in natural basalts during melt-peridotite interaction

Although Mg and Fe isotopes are frequently used to trace the source heterogeneity of basalts (e.g., recycled carbonated materials or eclogite in the mantle), they can also be fractionated by geological processes during magma migration from the source to the surface. Recently, the light Mg and heavy Fe isotopic compositions observed in high-Ti lunar basalts have been attributed to the kinetic fractionation that occurs during interaction between the melt and ambient harzburgite cumulates. Due to the thick lithospheric mantle, the melt-peridotite interaction is frequently invoked to explain the chemical variations observed in continental basalts on Earth. However, the isotope fractionation behavior of MgFe isotopes during this process has not been well addressed. Here we present MgFe isotopic compositions of alkaline basalts from the Nanjing basaltic field in eastern China, alongside the peridotite xenoliths hosted within these basalts, in order to assess the potential isotopic fractionation under this specific condition. Previous petrographic evidence suggests that these peridotite xenoliths had undergone significant interaction with the hosting basalts. Peridotite xenoliths exhibit homogeneous, normal-mantle-like δ26Mg (from −0.243 ‰ to −0.281 ‰), but they have highly variable, low δ57Fe (from 0.156 ‰ to −0.057 ‰), indicating that melting reaction has not altered the Mg isotopic compositions but has modified the Fe isotopic compositions of these xenoliths. The hosting basaltic rocks exhibit heavy Fe isotopic compositions (δ57Fe = 0.218 ‰–0.366 ‰) and extremely light Mg isotopic compositions (δ26Mg = −0.641 ‰−−0.389 ‰). The δ26Mg values of these xenolith-rich basalts are well correlated with their δ57Fe values, Hf isotopes, and certain elemental ratios (e.g., Lu/Hf, La/Sm, Ca/Al, and Na/Ti). These correlations can be explained by mixing between the host basaltic magma and a type of melt derived from peridotite xenolith. These xenolith-derived melts have significantly lighter Mg and heavier Fe isotopic compositions than the normal mantle. Given their extremely low Mg/Fe ratios resulting from near-solidus melting, they are not in Mg or Fe equilibrium with the residual solid. Consequently, kinetic MgFe isotope fractionation is inferred to occur during the interaction between the melt and peridotite. Therefore, our results serve as a representative sample for characterizing MgFe isotope fractionation via chemical diffusion in natural systems, and they highlight the significant influence of melt-peridotite interaction on the heterogeneity of MgFe isotopes in those xenolith-rich basalts (characterized by low melt/olivine ratios).

Supervised multiple kernel learning approaches for multi-omics data integration

BackgroundAdvances in high-throughput technologies have originated an ever-increasing availability of omics datasets. The integration of multiple heterogeneous data sources is currently an issue for biology and bioinformatics. Multiple kernel learning (MKL) has shown to be a flexible and valid approach to consider the diverse nature of multi-omics inputs, despite being an underused tool in genomic data mining.ResultsWe provide novel MKL approaches based on different kernel fusion strategies. To learn from the meta-kernel of input kernels, we adapted unsupervised integration algorithms for supervised tasks with support vector machines. We also tested deep learning architectures for kernel fusion and classification. The results show that MKL-based models can outperform more complex, state-of-the-art, supervised multi-omics integrative approaches.ConclusionMultiple kernel learning offers a natural framework for predictive models in multi-omics data. It proved to provide a fast and reliable solution that can compete with and outperform more complex architectures. Our results offer a direction for bio-data mining research, biomarker discovery and further development of methods for heterogeneous data integration.

Multi-Model Fusion Demand Forecasting Framework Based on Attention Mechanism

The accuracy of demand forecasting is critical for supply chain management and strategic business decisions. However, as data volumes grow and demand patterns become increasingly complex, traditional forecasting methods encounter significant challenges in processing intricate multi-dimensional data and achieving a satisfactory predictive accuracy. To address these challenges, this paper proposed an end-to-end multi-model demand forecasting framework based on attention mechanisms. The framework employs a dual attention mechanism to dynamically extract features from both the temporal and product dimensions, while integrating conditional information captured through convolutional neural networks, thereby enhancing its ability to model complex demand patterns. Additionally, a channel attention mechanism is introduced to perform the weighted fusion of outputs from multiple predictive models, thereby overcoming the limitations of single-model approaches and improving adaptability to varying demand patterns across diverse scenarios. The experimental results demonstrate that the proposed method outperforms conventional approaches across several evaluation metrics, achieving a 42% reduction in Mean Squared Error (MSE) compared to the baseline model. This notable improvement enhances both the accuracy and stability of demand forecasting. The framework offers valuable insights for addressing large-scale and complex demand patterns, providing guidance for precise decision-making and resource optimization within supply chain management. Future research will concentrate on further enhancing the model’s generalization capability to manage missing data and demand fluctuations. Additionally, efforts will focus on integrating diverse heterogeneous data sources to assess its performance in various practical scenarios, ultimately improving the model’s accuracy and flexibility.

Abel transform of laser-induced plasma image: Overcoming challenges of noisy and insufficient data for reliable reconstruction

Reconstructing spatial distribution of emissivity is important for study of heterogeneous plasma sources. However, often used Radon and Abel transforms tend to amplify noise in the data. The small size and short lifetime of laser-induced plasma result in short exposures for signal acquisition, and, consequently limited number of points along the source available for measurements and noisy signals. We compared 14 Abel transform algorithms and various noise processing methods, focusing on both limited data points and high noise levels. All algorithms were presented in matrix form to ensure fair comparison, and we proposed a metric based on the largest singular value of matrix operator for error assessment. An error analysis is performed to show the contribution of approximation and noise errors to the overall reconstruction error. Our results indicate that at higher noise levels, the choice of noise reduction method is more critical than the inversion algorithm itself, with regularization proving to be the most effective for noise control. Given the minimal impact of algorithm choice on the final result, simple and easy-to-implement methods like “Onion Peeling” are recommended to be combined with regularization. Furthermore, transforms with regularization were the most stable for reconstructing complex profiles. Summarizing the results for all considered cases, “Hankel-Fourier” and “Cubic Spline” are the most accurate for noiseless data with limited number of spatial points, while for cases with a large amount of data, “Polynomial”, “Piessens-Verbaeten”, and “Minerbo-Levy” are better, especially for noisy data.

Time-Dependent Evolution and Source Heterogeneities of Ocean Island Basalts from a Weak Plume, São Jorge, Azores

Abstract The geochemical composition of Ocean Island Basalts (OIBs) from the Azores reflects the spatial distribution, shape, and temporal evolution of small-scale geochemical heterogeneities within their mantle plume source. Here, we investigate the time-related evolution of volcanism at São Jorge Island, Central Azores. New field observations, a magnetic survey, 40Ar/39Ar and 14C ages and geochemical data indicate that the fissural volcanic activity at São Jorge produced at least four main mafic volcanic complexes (V. C.). The oldest V. C., São João, produced the thickest lava piles at ca. 1.3 Ma in the eastern part of the island. After a period of quiescence, the Serra do Topo V. C. was produced at ca. 0.8-0.5 Ma in the central part of the island. The Rosais V. C. was emplaced between ca. 0.4 and 0.1 Ma on the entire island. Finally, the Holocene Manadas V.C. volcanism became active in the western part of the island and includes three historic eruptions (1580, 1808 and 1964 CE). Magmas were formed at low melting degrees from a peridotitic mantle with possible minor contributions of recycled components. Olivine compositions and whole-rock trace element ratios discard a significant contribution from pyroxenitic source rocks. Melting temperatures (ca. 1420-1480 °C) were slightly higher than those of the ambient upper mantle. The four V.C. are characterized by distinct geochemical compositions in terms of incompatible trace elements and Sr-Nd-Pb isotopic ratios. The oldest V.C., São João, is characterized by Pb isotopic compositions (e.g., markedly negative Δ7/4 and 8/4 values) plotting well below the Northern Hemisphere Reference Line (NHRL). The Upper Pleistocene (Rosais V.C.) lavas from the north-western cliffs have compositions similar to enriched mantle (EM) end-member basalts (e.g., high 207Pb/204Pb at moderate 206Pb/204Pb; high Ba/Nb), which are rare among northern hemisphere OIBs. Finally, high 206Pb/204Pb (up to 20), reflecting contribution from a HIMU-type component characterizes the Holocene Manadas lavas and is occasionally found in lavas from other volcanic complexes from 1.3 to 0.1 Ma. These findings indicate that magmas from São Jorge and the nearby Central Azores islands were sourced from a strongly heterogeneous mantle plume, which displayed localized filaments of heterogeneous material that were rapidly exhausted (in ca. 0.2 Ma). The dominant component at São Jorge and in the Central Azores in general appears to be the HIMU-type endmember, which instead is not significant in the Eastern Azores. Possibly, the Central and Eastern Azores were produced by distinct branches of the Azores mantle plume, as would also be consistent with previous tomographic studies.

A spatial source-oriented and probability-based risk-assessment framework for heavy metal and PAH contamination of urban soils in Guangzhou, China

The identification and quantification of high-risk hotspots for soils contaminated by heavy metals (HMs) and polycyclic aromatic hydrocarbons (PAHs) remains a challenge due to their various sources and heterogeneous sink properties in urban soil systems. In this study of 221 soil samples from Guangzhou, China, a novel framework combining Bivariate local Moran’s I (BLMI), positive matrix factorization (PMF), human health risk (HHR) assessment, Monte Carlo simulation (MCS), and a newly developed spatial risk model were proposed to conduct probabilistic source-oriented HHR assessment, high-risk hotspot quantification, and risk formation mechanism elaboration. Study results indicate that traffic emissions are the largest contributor of HMs (47.6 %) and PAHs (40.2 %), but not always the largest contributor of HHR. Agricultural or urban green-space management activities of HM, and mixed source of PAH, are the largest contributors of non-carcinogenic risk (NCR, 48.7 % and 51.1 %, respectively), while mixed source of HM and traffic emissions of PAH are the largest contributors of carcinogenic risk (CR, 53.9 % and 71.2 %, respectively). The probability of risk exceeding safe threshold levels is < 5.0 % for NCR and > 90.0 % for CR. High-risk hotspots were identified in the mid-west and south of the city, making up 15.0 % of the total Guangzhou area. Risk mechanisms were deduced from the spatial heterogeneity and inter-dependence of emission sources and soil sink, based on source–sink theory. Our findings provide a new framework for precisely identifying risk sources and target areas, thereby alleviating HHR associated with co-occurring HMs and PAHs in urban soil systems.

Enhancing Compliance and Governance through Data Consistency and Rationalization for Effective Risk Mitigation in Health Care

In the era of big data, you absolutely must have consistency across your different data systems to make effective decision-making, be compliance, enhanced governance and security and also mitigation of risks,, and even have system integrity. The data consistency and how rationalization techniques help overcome challenges when working with heterogeneous data sources, is what this paper explores. Rationalization allows businesses to keep their data accurate and consistent by standardizing data formats, reducing duplications, measuring the data growth, categories the data sources, farmats, and aligning conflicting information. We compare the different methods of normalization, data deduplication, Master Data Management (MDM) and data integration framework. We also delve into the part that automation, artificial intelligence and machine learning play in optimizing these processes and provide scalable solutions that can potentially simplify the complexity of modern-day data environments. Practical implications from rationalization effort case studies are shown from sectors like, healthcare, finance, and e-commerce. Our specific findings show that long-term data consistency requires a systematic approach that employs technology and strong governance. In the end, this paper gives a blueprint for how to exploit rationalization methods to exploit the capabilities of data-driven insights fully.

Multimodal feature integration network for lithology identification from point cloud data

Accurate lithology identification from outcrop surfaces is crucial for interpreting geological 3D data. However, challenges arise due to factors such as severe weathering and vegetation coverage, which hinder achieving ideal identification results with both accuracy and efficiency. The integration of 3D point cloud technology and deep learning methodologies presents a promising solution to address these challenges. In this study, we propose a novel multimodal feature integration network designed to distinguish various rock types from point clouds. Our network incorporates a multimodal feature integration block equipped with multiple attention mechanisms to extract representative deep features, along with a hierarchical feature separation block to leverage these features for precise segmentation of points corresponding to different lithologies. Furthermore, we introduce a specialized loss function tailored for rock type identification to enhance network training. Through experiments involving point cloud sampling strategies and loss function evaluation, we identify the optimal network configuration. Comparative analyses against baseline methods demonstrate the superiority of our proposed network across diverse study areas reconstructed from UAV images and laser scanner data, exhibiting improved visual appearance and metric values (Accuracy = 0.978, mean Accuracy = 0.895, mean IoU = 0.857). These findings underscore the efficacy of the multimodal feature integration network as a promising approach for lithology identification tasks in various digital outcrop models derived from heterogeneous data sources.

A Multimodal conceptual framework to achieve automated software evolution for context-rich intelligent applications

AbstractWhile AI is extensively transforming Software Engineering (SE) fields, SE is still in need of a framework to consider overall all phases to facilitate Automated Software Evolution (ASEv), particularly for intelligent applications that are context-rich instead of conquering each division independently. Its complexity comes from the intricacy of the intelligent applications, the heterogeneity of the data sources, and the constant changes in the context. This study proposes a conceptual framework for achieving automated software evolution, emphasizing the importance of multimodality learning. A Selective Sequential Scope Model (3 S) model is developed based on the conceptual framework, and it can be used to categorize existing and future research when it covers different SE phases and multimodal learning tasks. This research is a preliminary step toward the blueprint of a higher-level ASEv. The proposed conceptual framework can act as a practical guideline for practitioners to prepare themselves for diving into this area. Although the study is about intelligent applications, the framework and analysis methods may be adapted for other types of software as AI brings more intelligence into their life cycles.

Federated Learning : A Paradigm Shift in Collaborative Machine Learning

Federated learning (FL) has emerged as an exceptionally promising method within the realm of machine learning, enabling multiple entities to jointly train a global model while maintaining decentralized data. This paper presents a comprehensive review of federated learning methodologies, applications, and challenges. We begin by elucidating the fundamental concepts underlying FL, including federated optimization algorithms, communication protocols, and privacy-preserving techniques. Subsequently, we delve into various domains where FL has found significant traction, examples include healthcare, finance, and the Internet of Things (IoT), showcasing successful deployments and innovative strategies. Furthermore, we discuss the inherent challenges associated with federated learning, such as communication overhead, heterogeneity of data sources, and privacy concerns, and explore state- of-the-art solutions proposed in literature. Finally, we outline future research directions in federated learning, including advancements in privacy-preserving techniques, scalability improvements, and extension of FL to emerging domains. This thorough examination provides a valuable asset for researchers, practitioners, and policymakers keen on grasping the panorama of federated learning and its ramifications for collaborative machine learning in dispersed settings.