Spatial Data Mining Research Articles

Spatial data mining-focused research on image processing techniques for remotely sensed images

ABSTRACT Spatial data mining is an important approach for collecting useful data from big datasets, especially remotely sensed images. This study tackles issues in environmental monitoring and management using sophisticated image processing. The Horse Herd Optimization-based VGG19 (HHO-VGG19) is proposed to improve land cover classification, recognition of objects, detection of changes, and detection of anomalies. The study used the BCDD dataset, which was scaled to 512 × 512 pixels, then applied Z-score normalization and extracted features using Principal Component Analysis (PCA). The VGG19 architecture was enhanced by utilizing Horse Herd Optimization to enhance image classification efficiency. The HHO-VGG19 model surpasses conventional techniques, with F1-score of 92%, a recall of 94%, an accuracy of 98.5%, and a 30-second execution time reduction. The findings indicate the efficiency of integrating sophisticated image processing with spatial data mining, giving an effective tool for remote sensing image processing in environmental uses including tracking ecosystems and handling of natural resources.

The degree of soil arsenic background enrichment by carbonate weathering is mainly controlled by climate in large spatial scale

Spatial distribution of soil arsenic (As) is heterogeneous. Making clear the dominant factor(s) controlling its spatial variation contributes to the differentiation of its natural background from anthropogenic pollution. Recent studies have found that the “high background” of soil heavy metals may be induced by the process of carbonate weathering. However, the extent, process, and factors that controls the degree of this enrichment, is still unclear, especially in large spatial scales. In this study, this problem is to be revealed by a compilation and spatial data mining of soil geochemical data for Australia, North America, Europe, and China with a collection and chemical analysis of bedrock and soil profiles in a typical region. The results indicated that the process of carbonate weathering strengthened the As enrichment in the soil; therefore, the carbonate rock underlaid regions could be distinguished from those of other bedrocks in the soil As maps. In the southwest region of China, an area larger than 12,000 km2 exceeding the intervention value of 100 mg/kg could be produced according to the national standard. Iron minerals play an important role for this enrichment as they serve as carriers for As during the stages of both carbonate leaching and the decomposition of silicate minerals. Geodetector results indicated a significant dependence on climate for the degree of enrichment, and both higher temperature and precipitation were necessary, which caused a decreasing trend of soil As on carbonate region from the warm and humid south to the cold and dry north in China. The dependence of soil As concentration on the mean average precipitation (MAP) and mean average temperature (MAT) was formulated using the data of the wider geographical regions across the world, and a theoretically predicted map has been produced to show the extent caused by this causation.

Aplicación de la minería de datos espacial basado en técnicas de agrupamiento para detectar los niveles de congestionamiento del tráfico vehicular en la ciudad de Trujillo

Aplicación de la minería de datos espacial basado en técnicas de agrupamiento para detectar los niveles de congestionamiento del tráfico vehicular en la ciudad de Trujillo

Spatial Patterns of Violence Against Women and Children using Geographic Information System and Density-Based Clustering Algorithm

This paper underscores the importance of the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm in data mining. In this research, data mining clustering methods were applied to investigate consummated felonies related to the "Anti-Violence Against Women and Their Children Act of 2004 - RA 9262" from 2018 to 2023. The criminal data processed from the Police Regional Office 6 of the Philippine National Police encompassed 248 attributes reflecting cases over the specified period. The significance of this study lies in its utilization of ArcGIS Pro software to process the provided data through clustering techniques, presenting a robust approach for detecting criminal activities and recognizing patterns to aid law enforcement in crime reduction efforts. Spatial data mining proves practical when dealing with geographic crime datasets, facilitating the analysis of large volumes of crime data. The DBSCAN algorithm was employed to cluster crime incidents centered on predefined events, with the resultant clusters used to identify hotspots. These clustering outcomes are then visualized using GIS, enabling real-time mapping of crime distribution for law enforcement agencies to comprehend and engage with effectively. The outcomes empower stakeholders to devise interventions tailored to specific locations, thereby contributing to a safer environment for women and children. The study illuminates the localized analysis of crime distribution, offering insights into the interconnected factors influencing criminal incidents and providing a framework for crafting targeted and efficient strategies for crime prevention, thereby enriching the broader dialogue on crime management and public safety.

Identifying a good business location using prescriptive analytics: Restaurant location recommendation based on spatial data mining

This study proposes a new prescriptive analytics method that aims to provide decision-makers with a systematic and objective approach to identify suitable locations, considering the spatial distribution of different types of restaurants. The method comprises of two main components: spatial co-location pattern mining and locationGCN, where locationGCN is based on graph convolutional network (GCN). The spatial co-location pattern mining is utilized to capture the spatial correlation of specific restaurant to determine the candidate location selection range. The locationGCN is designed to further screen out final suitable location ranges for the specific restaurant type. A case study using restaurant data from Xiamen Island collected from Dianping.com is conducted. The empirical results demonstrate that the algorithm achieves an accuracy of 74.88%, precision of 63.59%, and recall of 77.48%. Results indicate that the proposed approach can provide suitable location recommendations for specific types of restaurants based on existing restaurant distribution information.

Machine Learning Approaches in Spatial Data Mining

This review paper surveys the integration of machine learning techniques in spatial data mining, a crucial intersection of geographic information systems and data mining. It examines the application of various machine learning algorithms such as classification, regression, clustering, and deep learning in spatial data analysis. The paper discusses challenges like data preprocessing, feature selection, and model interpretability, alongside recent advancements including spatial-temporal analysis and heterogeneous data integration. Through critical analysis of existing literature, it identifies trends, methodologies, and future research directions. Practical implications and applications across domains like urban planning, environmental monitoring, and epidemiology are explored. As a comprehensive resource, this review facilitates understanding and utilization of machine learning approaches for extracting insights from spatial data, benefiting researchers, practitioners, and policymakers alike.

Navigating the landscapes of spatial transcriptomics: How computational methods guide the way.

Spatially resolved transcriptomics has been dramatically transforming biological and medical research in various fields. It enables transcriptome profiling at single-cell, multi-cellular, or sub-cellular resolution, while retaining the information of geometric localizations of cells in complex tissues. The coupling of cell spatial information and its molecular characteristics generates a novel multi-modal high-throughput data source, which poses new challenges for the development of analytical methods for data-mining. Spatial transcriptomic data are often highly complex, noisy, and biased, presenting a series of difficulties, many unresolved, for data analysis and generation of biological insights. In addition, to keep pace with the ever-evolving spatial transcriptomic experimental technologies, the existing analytical theories and tools need to be updated and reformed accordingly. In this review, we provide an overview and discussion of the current computational approaches for mining of spatial transcriptomics data. Future directions and perspectives of methodology design are proposed to stimulate further discussions and advances in new analytical models and algorithms. This article is categorized under: RNA Methods > RNA Analyses in Cells RNA Evolution and Genomics > Computational Analyses of RNA RNA Export and Localization > RNA Localization.

Measuring the similarity between shapes of buildings using graph edit distance

ABSTRACT Building data play a crucial role in geographic information science, and building shape similarity can be used for spatial inquiry, cartographic generalization, map updating, data quality assessment, and other spatial data mining applications. However, conventional shape representation and measurement methods use geometric and statistical measures and may not effectively capture the subtle differences among building shapes. Geospatial artificial intelligence methods often overlook nuanced differences in building contours, posing a challenge when precise fine-grained measurements are required. Furthermore, they have low interpretability. This study proposes a graph edit distance-based method for measuring the similarity between building shapes. It uses geometrics characteristics to construct a graph, matches and associates similar regional features between two building contours, and defines edit costs based on geometric and positional differences that align with the Gestalt principle of visual similarity. It converts the optimal edit path problem to the maximum weighted cluster solution problem and utilizes a heuristic approach to achieve the exact optimal solution within a moderately suitable timeframe. The experimental results demonstrated that the proposed method identifies detailed differences between building contours, providing accurate similarity measurement with high interpretability.

The Logical Mechanism and Practical Progress of Digital Economy Contributing to Rural Revitalization in the Context of Big Data

Abstract This paper proposes spatial data mining and clustering analysis techniques based on big data and uses four distance similarity calculation methods, namely, Euclidean distance, Manhattan distance, Ming’s distance, and Ma’s distance, to constrain spatial clustering similarity metrics. Combine the K-means algorithm and FCM clustering algorithm to calculate the center value of fuzzy clustering, use the additive synthesis method to solve the combination weights, and construct the combination weighted distance, fuzzy C-mean algorithm model. Research on the mechanism and path of digital economy-enabled rural revitalization through the C-FCM model, and propose three correlation hypotheses for this paper. Empirical experiments are conducted to verify the authenticity of the hypotheses proposed in this paper and the use effect of the constructed model. The experimental results show that the Moran index range of digital economy and rural revitalization is [0.06, 0.12], [0.05, 0.16], respectively, and there is a significant aggregation phenomenon of the two under the geographic matrix. In the analysis of the spatial spillover effect, the values of spatial autoregressive coefficients of rural industrial revitalization are all >0, with significance p>0.05. The digital economy has a positive effect on rural revitalization in neighboring areas through the spatial spillover effect. In the clustering sample extraction test, the accuracy is most stable when the number of samples is 2200, and the accuracy rate is as high as 79.66%.

Optimal electric vehicle scheduling in unbalanced distribution system by spatial and temporal data mining and bi‐level particle swarm optimisation

AbstractThis research proposes a new comprehensive methodology aimed at optimally managing electric vehicle (EV) charging in an unbalanced distribution system with consideration of grid capacity and voltage quality constraints as well as the stochastic driving and charging behaviour of EV users. A data mining algorithm is designed to generate suitable spatial and temporal EV and charger input data for EV scheduling that is decomposed into two subproblems. The lower‐level subproblem identifies the maximum load at each node and time and the upper‐level subproblem allocates charging slots for each EV being charged. Both subproblems are solved by developed particle swarm optimisation (PSO) algorithms. The effectiveness and robustness of the algorithms have been thoroughly validated by conducting rigorous tests on a modified IEEE 37‐bus distribution system with different EV penetration scenarios. The test results have confirmed the algorithms' performance in accommodating the increasing load associated with EV charging and successfully improving the system performance by maximising the system load factor, minimising load unbalance, and reducing the system power loss. All operational constraints on node voltages, and distribution transformer and feeder capacities of the existing power distribution infrastructure are fully respected while maintaining high user satisfaction represented by the average state of charge (SoC) of all EVs.

Graph based embedding learning of trajectory data for transportation mode recognition by fusing sequence and dependency relations

As an important task in spatial data mining, trajectory transportation mode recognition can reflect various individual behaviors and traveling patterns in urban space. As trajectory is essentially a sequence, many scholars use the sequence inference models to mine the information in trajectory data. However, such methods often ignored the spatial correlation between trajectory points and implemented the evaluation based only on representative feature statistics selected in the trajectory data preprocessing stage, thus have difficulties in acquiring high-order traveling pattern features. In this study, we propose a novel ensemble recognition method for representing trajectory data with the graph structure based on sequence and dependency relations. This method integrates the sequence of trajectory points and the correlation between characteristic points of a travel path into a fused graph convolutional network to obtain semantic feature information at multiple levels. We validate our proposed method with experiments on the trajectory benchmark dataset from the Microsoft GeoLife project. The results demonstrated that our proposed graph network outperforms other baseline methods in the transportation mode recognition task of trajectories. This method can help to discover the movement patterns of urban residents, and further provide effective assistance for the management of cities.

Does partition matter? A new approach to modeling land use change

Cellular automata (CA) is a classical method for studying land use change. However, homogeneous transformation rules have commonly been used to conduct simulations in the past, and these rules seldom consider the spatial heterogeneity of geographic elements. Therefore, in this study, we incorporated spatial-temporal heterogeneity transformation rules into the CA framework based on spatial data mining. A model that couples self-organizing maps (SOM), hierarchical clustering (HC), and patch generation land use simulation (PLUS) was proposed; it is called the SOM-HC-PLUS model. This model considers the difference in local-area driving factors, and therefor it not only determines the optimal partition scheme automatically, but it also measures the contribution of partition driving factors. The Jinpu New Area in Dalian, China, was used to test the validity of the model by comparing the traditional PLUS model and the administrative division (AD)-PLUS model based on AD partition. The results showed that the partition scheme of the SOM-HC-PLUS model was reasonable and credible. Further, compared with other models, this model showed higher simulation accuracy and a more realistic land use distribution pattern. The driving factors showed significant differences in the overall and regional intensity. Moreover, the importance of natural environmental conditions, represented by elevation factors, in the expansion of artificial surfaces increased significantly. By 2030, artificial surfaces were projected to increase significantly through the conversion of cultivated lands. The sustainable development scenario showed a more compact patch layout and exhibited better protection of grasslands and forests than the historical development scenario. In summary, this study proposed a mixture CA model based on the idea of geographic partition, one that proved the reliability of the SOM-HC-PLUS model to conduct spatial-temporal heterogeneity studies on land use partition. It provides the possibility to explore patterns of regional land use changes over multiple periods, and can assist in urban planning and management and promoting sustainable development.

Combining the Back Propagation Neural Network and Particle Swarm Optimization Algorithm for Lithological Mapping in North China

Lithological mapping is a crucial tool for exploring minerals, reconstructing geological formations, and interpreting geological evolution. The study aimed to investigate the application of the back propagation neural network (BPNN) and particle swarm optimization (PSO) algorithm in lithological mapping. The study area is the Beiliutumiao map-sheet (No. K49E011021) in Inner Mongolia, China. This area was divided into two parts, with the left side used for training and the right side used for validation. Fifteen geological relevant factors, including geochemistry (1:200,000-scale) and geophysics (1:50,000-scale), were used as predictor variables. Taking one lithology as an example, the lithological binary mapping method was introduced in detail, and then the complete lithology was mapped. The model was compared with commonly used spatial data mining methods using the E-measure, S-measure, and Weighted F-measure values. In diorite testing, the accuracy and kappa of the optimized model were 92.11% and 0.81, respectively. The validation results showed that our method outperformed the traditional BPNN and weights-of-evidence approaches. In the extension of the complete lithological mapping, the accuracy, recall, and F1-score were 82.66%, 74.54%, and 0.76, respectively. Thus, the proposed method is useful for predicting the distribution of one lithology and completing the whole lithological mapping at a fine scale. In addition, the trained network can be extended to an adjacent area with similar lithological features.

An Analysis of Text Mining in Big Data

The practice of extracting hidden predictive information from a database and structuring it for later use is known as data mining. Web mining, text mining, sequence mining, graph mining, temporal data mining, spatial data mining, distributed data mining, and multimedia mining are some of the diverse domains in data mining. Financial data analysis, the retail and telecommunications sectors, science and engineering, and intrusion detection and prevention are a few fields where data mining is used. Worldwide, big data analytics is a burgeoning research field in computer science and many other fields. We covered text mining methods and their uses in this paper. Text mining is becoming a significant topic of study. Text mining is the process of automatically finding new, previously undiscovered information using a computer.

Understanding children's cycling route selection through spatial trajectory data mining

A sustainable alternative transportation mode to address growing transportation and environmental stress is cycling, which is eco-friendly and healthy for humans. Improving the quality of the bicycling experience is crucial for increasing bicycle use. Good bicycling experience is more critical for child bicyclists because they are less experienced and need more space for error. Therefore, a scientific assessment of child bicyclist perception of selected route safety, comfort, and environment is of great interest. Finding an effective way to learn child bicyclist behavior and help them reduce cycling risk is necessary. In response to this need, we utilize a data mining model to develop a methodology for measuring children's bicycling route safety conditions by evaluating multiple road safety-related features. The proposed method uses a set of route features representing the situation of street environments extracted from state data and first-hand children’s bicycle trajectory data collected using Global Positioning Systems (GPS) from volunteer children bicyclists. A random forest (RF), a well-known classifier, is adopted to predict child bicyclists' behavior. We extract the different route segments between children’s selected routes and the shortest path to learn the child bicyclists' behavior and use the selected best features to interpret their changing cycling behavior. The result shows that children bicyclists' behavior could be analyzed by giving trajectory and nearby road safety situation data. Our model achieves a promising accuracy with an average rate of 92% over multiple scenarios, demonstrating the proposed method's feasibility and the effectiveness of selected features. In addition, we compare our feature effectiveness with the state's generated road safety score data to evaluate the feature robustness. Our features outperform the safety score feature with an average of 10% improvement in prediction accuracy. Furthermore, our method proposes a model framework that can be applied to different study regions and adult bicycling behavior learning.

DASS Good: Explainable Data Mining of Spatial Cohort Data.

Developing applicable clinical machine learning models is a difficult task when the data includes spatial information, for example, radiation dose distributions across adjacent organs at risk. We describe the co-design of a modeling system, DASS, to support the hybrid human-machine development and validation of predictive models for estimating long-term toxicities related to radiotherapy doses in head and neck cancer patients. Developed in collaboration with domain experts in oncology and data mining, DASS incorporates human-in-the-loop visual steering, spatial data, and explainable AI to augment domain knowledge with automatic data mining. We demonstrate DASS with the development of two practical clinical stratification models and report feedback from domain experts. Finally, we describe the design lessons learned from this collaborative experience.

Optimization of Energy-Efficient Cluster Head Selection Algorithm for Internet of Things in Wireless Sensor Networks

The Internet of Things (IoT) now uses the Wireless Sensor Network (WSN) as a platform to sense and communicate data. The increase in the number of embedded and interconnected devices on the Internet has resulted in a need for software solutions to manage them proficiently in an elegant and scalable manner. Also, these devices can generate massive amounts of data, resulting in a classic Big Data problem that must be stored and processed. Large volumes of information have to be produced by using IoT applications, thus raising two major issues in big data analytics. To ensure an efficient form of mining of both spatial and temporal data, a sensed sample has to be collected. So for this work, a new strategy to remove redundancy has been proposed. This classifies all forms of collected data to be either relevant or irrelevant in choosing suitable information even before they are forwarded to the base station or the cluster head. A Low-Energy Adaptive Clustering Hierarchy (LEACH) is a cluster-based routing protocol that uses cluster formation. The LEACH chooses one head from the network sensor nodes, such as the Cluster Head (CH), to rotate the role to a new distributed energy load. The CHs were chosen randomly with the possibility of all CHs being concentrated in one locality. The primary idea behind such dynamic clustering was them resulted in more overheads due to changes in the CH and advertisements. Therefore, the LEACH was not suitable for large networks. Here, Particle Swarm Optimization (PSO) and River Formation Dynamics are used to optimize the CH selection (RFD). The results proved that the proposed method to have performed better compared to other methods.

Clustering objects for spatial data mining: a comparative study

Spatial data mining (SDM) is searching important relationships and characteristics that can clearly exist in spatial databases. This content aims to compare object clustering algorithms for spatial data mining, before identifying the most efficient algorithm. To this end, this paper compare k-means, Partionning Around Medoids (PAM) and Clustering Large Applications based on RANdomized Search (CLARANS) algorithms based on computing time. Experimental results indicate that, CLARANS is very efficient and effective.

Water Pollution Increases the Risk of Chytridiomycosis in Mexican Amphibians

Chytridiomycosis is affecting amphibians worldwide, causing the decline and extinction of several amphibian populations. The disease is caused by the fungus Batrachochytrium dendrobatidis (Bd), a multihost pathogen living in freshwater habitats. While several environmental factors have been associated with the prevalence of Bd and its virulence, the effects of water quality on the pathogen are not clear yet. Some evidence suggests that water pollution may reduce amphibians' immune response and increase prevalence of Bd. To explore this hypothesis, we analyzed the relationship between water quality and the presence of Bd by using spatial data mining of 150 geolocations of Bd in amphibians from 9 families where Bd positive specimens have been previously reported, and water quality in 4,202 lentic and lotic water bodies in Mexico from 2010 to 2021. Our model showed that in the 3 main families where Bd was recorded, its presence is high in locations with low water quality, i.e., water polluted likely contaminated with urban and industrial waste. Using this model, we inferred areas suitable for Bd in Mexico; mainly in poorly studied areas along the gulf and on the pacific slope. We further argue that actions to reduce water pollution should become an integral part of public policies to prevent the spread of Bd and protect amphibians from this deadly pathogen.

A chorem-based approach to visualizing COVID-19-related data

ABSTRACT Due to the increasing availability and openness of data from many countries worldwide, visualizations of the COVID-19 pandemic that aim to track the virus over time and across geographic areas are multiplying; geovisualization is proving to be an effective solution to better understand the spread of SARS-CoV-2. However, due to the resulting “infodemic,” many visualization approaches are sub-optimal. For example, direct projections of COVID-19-related data on display are unsound from a cartographic point of view; the results are often overcomplicated, leading to cognitive overload, which may in turn, lead to misinterpretations. Thus, developing effective geovisualization methods is necessary to improve epidemiologists’ and local authorities’ decision-making abilities. This study proposes a novel approach to geovisualizing COVID-19-related data based on chorems. Our objective was to define a chorematic map that could represent, through visual summaries extracted from the analysis of COVID-19-related data, the origin of the first cases and evolution of the virus in Algeria during the first months of the pandemic. Accordingly, we propose a geovisualization methodology based on spatial data mining techniques and GIS technology.