Multi-source Database Research Articles

Relationship between Residential Patterns and Socioeconomic Statuses Based on Multi-Source Spatial Data: A Case Study of Nanjing, China

The relationship between residential patterns and socioeconomic statuses highlights the complex interactions between the economic regime, welfare system, and neighborhood effects, which are crucial in urban inequality studies. With the diversification of the housing demand and supply system, the traditional analysis conducted separately from the ethnic or spatial segregation perspective fails to capture the rising inequalities and changing socio-spatial context. Taking Nanjing as an example, based on a multi-source database including the housing price, residential environmental quality, surrounding support facilities, and mobile phone user portrait data, this paper proposed a modified method for discovering the coupling relationship between residential patterns and socioeconomic statuses. It is found that socioeconomic status contributes to residential spatial aggregation and that the relationship between social and spatial dimensions of residential differentiation is tightly coupled and related. The lower socioeconomic strata were displaced to the periphery and the older urban core, while affluent inhabitants were more likely to settle voluntarily in segregated enclaves to isolate themselves from the general population through more flexible housing options. The heterogeneity of the urban socioeconomic dimension is primarily affected by consumption and occupational status, while housing prices mainly determine the divergence of spatial distribution.

Institutional Environment, Institutional Arrangements, and Risk Identification and Allocation in Public–Private Partnerships: A Multilevel Model Analysis Based on Data from 31 Provinces in China

As an important part of market-based reforms, the issue of “risk” has always been a part of the public–private partnership (PPP) debate, and the way in which risks are managed determines the sustainability of market-based reforms. This study systematically examines how the institutional environment and institutional arrangements affect the effectiveness of PPP risk identification and allocation. The study aims to establish a multi-source database of all publicized PPP projects in 31 provinces during the period of 2017–2021 in China, and it incorporates different levels of influencing factors, such as the institutional environment and institutional arrangements, into a framework. Through the application of a multilevel model, the impact of the political–institutional structure at the provincial level and institutional arrangements at the social capital and project levels on PPP risk identification and allocation is judged hierarchically. It is found that only social capital at the institutional environment level can directly and positively contribute to the effective identification and allocation of PPP risks, while both the degree of potential market competition and the degree of integration at the project level have a positive effect on its risk profile. In addition, there are cross-level moderating effects, with social capital, government transparency, and government–business relations positively improving risk allocation and identification through the degree of market competition, with no significant coordinating effect on the degree of contractual integration. This study goes beyond the existing one-dimensional risk profile analysis and risk typology to dismantle the “black box” of risk identification and allocation, which is a process of continuous negotiation, and to provide a sustainable governance mechanism for the market-oriented reform of public services in the institutional arena.

Joint diagnosis and validation of lightning risk for transmission lines by using multisource data

This study proposes an improved method for assessing lightning risk for typical 750-kV transmission lines in Northwest China. The method involves enhanced preprocessing of multisource parameters and a joint diagnosis of lightning risk; on the basis of the diagnosis results, a cause analysis is conducted to identify towers at high lightning risk. In this study, raw data on the following parameters were obtained and processed: line lightning withstand capability calculation parameters, lightning localization parameters, microtopography parameters, and distributed monitoring parameters. These data were then integrated into a comprehensive multisource database. Furthermore, this study established a model for calculating lightning withstand capability; the study considered the influence of altitude and optimized the classic electrogeometric model. The joint diagnosis method was validated using distributed monitoring data. Through this method, this study identified the characteristics of typical high-risk towers and the underlying factors contributing to the high lightning risk on transmission lines.

Assessment of Mycological Possibility Using Machine Learning Models for Effective Inclusion in Sustainable Forest Management

The integral role of wild fungi in ecosystems, including provisioning, regulating, cultural, and supporting services, is well recognized. However, quantifying and predicting wild mushroom yields is challenging due to spatial and temporal variability. In Mediterranean forests, climate-change-induced droughts further impact mushroom production. Fungal fruiting is influenced by factors such as climate, soil, topography, and forest structure. This study aims to quantify and predict the mycological potential of Lactarius deliciosus in sustainably managed Mediterranean pine forests using machine learning models. We utilize a long-term dataset of Lactarius deliciosus yields from 17 Pinus pinaster plots in Soria, Spain, integrating forest-derived structural data, NASA Landsat mission vegetation indices, and climatic data. The resulting multisource database facilitates the creation of a two-stage ‘mycological exploitability’ index, crucial for incorporating anticipated mycological production into sustainable forest management, in line with what is usually done for other uses such as timber or game. Various Machine Learning (ML) techniques, such as classification trees, random forest, linear and radial support vector machine, and neural networks, were employed to construct models for classification and prediction. The sample was always divided into training and validation sets (70-30%), while the differences were found in terms of Overall Accuracy (OA). Neural networks, incorporating critical variables like climatic data (precipitation in January and humidity in November), remote sensing indices (Enhanced Vegetation Index, Green Normalization Difference Vegetation Index), and structural forest variables (mean height, site index and basal area), produced the most accurate and unbiased models (OAtraining = 0.8398; OAvalidation = 0.7190). This research emphasizes the importance of considering a diverse array of ecosystem variables for quantifying wild mushroom yields and underscores the pivotal role of Artificial Intelligence (AI) tools and remotely sensed observations in modeling non-wood forest products. Integrating such models into sustainable forest management plans is crucial for recognizing the ecosystem services provided by them.

The Use of Human Amniotic Membrane (hAM) as a Treatment Strategy of Medication-Related Osteonecrosis of the Jaw (MRONJ): A Systematic Review and Meta-Analysis of the Literature.

Background and objectives: Although it is very uncommon, medication-induced osteonecrosis of the jaw (also known as MRONJ) can have serious consequences. Traditionally, this adverse event has been recognised in patients who were treated with bisphosphonate (BP) drugs. Nevertheless, in recent years, it has been established that individuals having treatment with various types of medications, such as a receptor activator of nuclear factor kappa-Β ligand inhibitor (denosumab) and antiangiogenic agents, have had the same issue. The purpose of this research is to determine if the application of human amniotic membrane (hAM) may be used as a therapy for MRONJ. Material and Methods: A multi-source database (MEDLINE, EMBASE, AMED, and CENTRAL) systematic search was performed. The major objective of this study is to obtain an understanding of the efficacy of hAM when it is employed as a treatment modality for MRONJ. The protocol of this review was registered in the INPLASY register under the number NPLASY202330010. Results: The authors were able to include a total of five studies for the quality analysis, whereas for the quantity evaluation, only four studies were eligible. A total of 91 patients were considered for the investigation. After treatment with human amniotic membrane (hAM), a recurrence of osteonecrosis was observed in n = 6 cases (8.8%). The combined efficacy of surgical therapy and the use of hAM resulted in an overall success rate of 91.2%. Intraoperative complications were only documented in one article, and they were mostly caused by the positioning of the hAM, which led to wound breakdown at the surgical site. Conclusions: Based on the small amount of data and low-quality research included in this study, using human amniotic membranes to treat MRONJ might represent a feasible option. Nevertheless, further studies with a wider patient population are required to understand the long-term impacts.

Development of a COVID-19 vaccine effectiveness and safety assessment system in Japan: The VENUS study.

There are currently no COVID-19 vaccine assessment systems in Japan that allow for the active surveillance of both vaccinated and unvaccinated persons. Herein, we describe the development of Japan's first COVID-19 vaccine effectiveness and safety assessment system with active surveillance capabilities. The Vaccine Effectiveness, Networking, and Universal Safety (VENUS) Study was developed as a multi-source database that links four data types at the individual resident level: Basic Resident Register (base population information), Vaccination Record System (vaccination-related information), Health Center Real-time Information-sharing System on COVID-19 (HER-SYS; information on COVID-19 occurrence), and health care claims data (information on diagnoses, hospitalizations, diagnostic tests, and treatments). These data were obtained from four municipalities. Individual residents were linked across the data types using five matching algorithms based on names, birth dates, and sex; the data were anonymized after linkage. To ascertain the viability of the VENUS Study's database for COVID-19 vaccine assessments, we examined the trends in COVID-19 vaccinations, COVID-19 cases, and polymerase chain reaction (PCR) test numbers. We also evaluated the linkage rates across the data types. Our multi-source database was able to monitor COVID-19 vaccinations, COVID-19 cases, and PCR test numbers throughout the pandemic. Using the five algorithms, the data linkage rates between the COVID-19 occurrence information in the HER-SYS and the Basic Resident Register ranged from 85·4% to 91·7%. If used judiciously with an understanding of each data source's characteristics, the VENUS Study can provide a viable data platform that facilitates active surveillance and comparative analyses for population-based research on COVID-19 vaccine effectiveness and safety in Japan.

Estimation of Aquifer Storativity Using 3D Geological Modeling and the Spatial Random Bagging Simulation Method: The Saskatchewan River Basin Case Study (Central Canada)

Hydrosystems in the Saskatchewan River Basin of the Canadian Prairies are subject to natural and socioeconomic pressures. Increasingly, these strong pressures are exacerbating problems of water resource accessibility and depletion. Unfortunately, the geometric heterogeneity of the aquifers and the presence of lithologically varied layers complicate groundwater flow studies, hydrodynamic characterization, and aquifer storativity calculations. Moreover, in recent hydrogeological studies, hydraulic conductivity has been the subject of much more research than storativity. It is in this context that the present research was conducted, to establish a 3D hydrostratigraphic model that highlights the geological (lithology, thickness, and depth) and hydrodynamic characteristics of the aquifer formations and proposes a new uncertainty framework for groundwater storage estimation. The general methodology is based on collecting and processing a very fragmentary and diverse multi-source database to develop the conceptual model. Data were harmonized and entered into a common database management system. A large quantity of geological information has been implemented in a 3D hydrostratigraphic model to establish the finest geometry of the SRB aquifers. Then, the different sources of uncertainty were controlled and considered in the modeling process by developing a randomized modeling system based on spatial random bagging simulation (SRBS). The results of the research show the following: Firstly, the distribution of aquifer levels is controlled by tectonic activity and erosion, which further suggests that most buried valleys on the Prairies have filled over time, likely during multiple glaciations in several depositional environments. Secondly, the geostatistical study allowed us to choose optimal interpolation variographic parameters. Finally, the final storativity maps of the different aquifer formations showed a huge potential of groundwater in SRB. The SRBS method allowed us to calculate the optimal storativity values for each mesh and to obtain a final storativity map for each formation. For example, for the Paskapoo Formation, the distribution grid of groundwater storage shows that the east part of the aquifer can store up to 5920 × 103 m3/voxel, whereas most areas of the west aquifer part can only store less than 750 × 103 m3/voxel. The maximum storativity was attributed to the Horseshoe Canyon Formation, which contains maximal geological reserves ranging from 107 to 111 × 109 m3. The main contribution of this research is the proposed 3D geological model with hydrogeological insights into the study area, as well as the use of a new statistical method to propagate the uncertainty over the modeling domain. The next step will focus on the hydrodynamic modeling of groundwater flow to better manage water resources in the Saskatchewan River Basin.

Impact of Clinical Practice Gaps on the Implementation of Personalized Medicine in Advanced Non-Small-Cell Lung Cancer.

Personalized medicine presents new opportunities for patients with cancer. However, many patients do not receive the most effective personalized treatments because of challenges associated with integrating predictive biomarker testing into clinical care. Patients are lost at various steps along the precision oncology pathway because of operational inefficiencies, limited understanding of biomarker strategies, inappropriate testing result usage, and access barriers. We examine the impact of various clinical practice gaps associated with diagnostic testing-informed personalized medicine strategies on the treatment of advanced non-small-cell lung cancer (aNSCLC). Using Diaceutics' Data Repository, a multisource database including commercial and Medicare claims and laboratory data from over 500,000 patients with non-small-cell lung cancer in the United States, we analyzed the number of patients with newly diagnosed aNSCLC who could have, but did not, benefit from a personalized treatment. The analysis focuses on the independent and cumulative impacts of gaps occurring during seven steps of the precision oncology pathway, from diagnosis to treatment. For every 1,000 patients in the study cohort, 497 (49.7%) are lost to precision oncology because of factors associated with getting biomarker test results. Among the 503 of 1,000 patients who did receive results from a biomarker test, 147 (29.2%) did not receive appropriate targeted treatments. Thus, approximately 64% of potentially eligible patients with aNSCLC are not benefiting from precision oncology therapies appropriate for their disease. Most patients with aNSCLC eligible for precision oncology treatments do not benefit from them because of clinical practice gaps. This finding is likely reflective of similar gaps in other cancer types. An increased understanding of the impact of each practice gap can inform strategies to improve the delivery of precision oncology, helping to fully realize the promise of personalized medicine.

Speaking of Opportunities: The Effect of Language on Entrepreneurial Alertness

Although entrepreneurial alertness is foundational in applied entrepreneurship studies, entrepreneurship scholars have rarely considered institutional factors that shape entrepreneurial alertness. We investigate the underlying mechanisms of entrepreneurial alertness with respect to one of the most salient, yet underexplored, aspects of informal institutions: the future-time-reference (FTR) in linguistic structures. Drawing upon imprinting theory, we theorize that entrepreneurs’ early language imprint persistently affects their alertness, and that both individual and environmental contextual factors may alter this effect. We test our model with two studies. Study 1 employs a multi-source database with 107,830 observations across 56 countries and Study 2 surveys 618 entrepreneurs in four countries. Our findings indicate that entrepreneurs in institutions with strong-FTR languages exhibit higher level of alertness. Further, this effect is intensified for older entrepreneurs, yet weakened for entrepreneurs with higher level of education or entrepreneurs who experience higher level of globalization in their home country. Our study contributes to a better understanding of language imprinting on entrepreneurial alertness.

A multisource database tracking the impact of the COVID-19 pandemic on the communities of Boston, MA, USA

A pandemic, like other disasters, changes how systems work. In order to support research on how the COVID-19 pandemic impacted the dynamics of a single metropolitan area and the communities therein, we developed and made publicly available a “data-support system” for the city of Boston. We actively gathered data from multiple administrative (e.g., 911 and 311 dispatches, building permits) and internet sources (e.g., Yelp, Craigslist), capturing aspects of housing and land use, crime and disorder, and commercial activity and institutions. All the data were linked spatially through BARI’s Geographical Infrastructure, enabling conjoint analysis. We curated the base records and aggregated them to construct ecometric measures (i.e., descriptors of a place) at various geographic scales, all of which were also published as part of the database. The datasets were published in an open repository, each accompanied by a detailed documentation of methods and variables. We anticipate updating the database annually to maintain the tracking of the records and associated measures.

Emerging weed resistance increases tillage intensity and greenhouse gas emissions in the US corn-soybean cropping system.

Tillage is a common agricultural practice that helps prepare the soil and remove weeds. However, it remains unknown how tillage intensity has evolved and its effect on net greenhouse gas (GHG) emissions. Here, using a process-based modelling approach with a multi-source database, we examined the change in tillage intensity across the US corn-soybean cropping systems during 1998-2016 and the impact of tillage intensity on soil GHG emissions. We found that tillage intensity first decreased and then, after 2008, increased, a trend that is strongly correlated with the adoption of herbicide-tolerant crops and emerging weed resistance. The GHG mitigation benefit (-5.5 ± 4.8 TgCO2e yr-1) of decreasing tillage intensity before 2008 has been more than offset by increased GHG emissions (13.8 ± 5.6 TgCO2e yr-1) due to tillage reintensification under growing pressure of weed resistance. As weed resistance persists or grows, tillage intensity is anticipated to continue rising, probably increasing GHG emissions. Our results imply that farmers' choices in managing herbicide resistance may help mitigate agricultural GHG emissions, underscoring the importance of an alternative strategy to control weeds.

Diosgenin From Dioscorea Nipponica Rhizoma Against Graves' Disease-On Network Pharmacology and Experimental Evaluation.

Dioscorea nipponica rhizoma (DNR) is commonly used for the cure of hyperthyroidism resulting from Graves’ disease (GD) or thyroid nodules. However, its therapeutic mechanism remains unclear. This study aimed to utilize network pharmacology integrated molecular docking and experimental verification to reveal the potential pharmacological mechanism of DNR against GD. First, the active componds of DNR were collected from the HERB database and a literature search was conducted. Then, according to multisource database, the predicted genes of DNR and GD were collected to generate networks. The analysis of protein–protein interaction and GO enrichment and KEGG pathway were employed to discover main mechanisms associated with therapeutic targets. Moreover, molecular docking simulation was applied in order to verify the interactions between the drug and target. Finally, our experiments validated the ameliorated effects of diosgenin, the main component of DNR, in terms of phosphorylation deactivation in IGF-1R, which in turn inhibited the phosphorylation and activation of PI3K-AKT and Rap1-MEK signaling pathways, promoting cell apoptosis and GD remission. Our present study provided a foundation for further investigation of the in-depth mechanisms of diosgenin in GD and will provide new scientific evidence for clinical application.

Multi-source data driven method for assessing the rock mass quality of a NATM tunnel face via hybrid ensemble learning models

Current assessments of rock mass quality of a NATM tunnel face are important in the practice of tunnel excavation. This study establishes a multi-source database and proposes a data driven method for the assessment. Thirteen multi-source variables describing the tunnel faces are considered as inputs, and the rock mass rating (RMR) values computed by the empirical formula are the target outputs. We adopted two meta machine learning models (classification and regression tree (CART) and multiple layers perceptron (MLP)) and two ensemble learning models (gradient boosting regression tree (GBRT)) and random forest (RF)) to capture the relationships between the inputs and outputs. The tree-structured Parzen estimator (TPE) algorithm is applied to automatically determine the optimized model hyper-parameters. The experimental results suggest that the proposed hybrid ensemble learning models (TPE-RF and TPE-GBRT) perform well at assessing rock mass quality. The feature importance ranks of the input variables are determined by a sensitivity analysis, which enhances the knowledge on assessing the rock mass quality of a tunnel face.

Comparing Cannabis Use Disorder in the General Population with Cannabis Treatment Seekers Using Multi-Source National Datasets: Who Receives Treatment?

Introduction: Given the increased prevalence of cannabis use in Ireland and increase in cannabis potency, this study aimed to estimate the size of the potential population in Ireland that may be in need of cannabis treatment and the percentage of people with cannabis use disorder (CUD) who actually access treatment. We also compared the profile of those with CUD in the general population to those who receive treatment for their cannabis use to explore whether certain subgroups are more or less likely to enter treatment. Method: This was a retrospective, multi-source database study. Data were obtained from (1) Ireland’s 2014/2015 national general population survey (GPS) on drug use and (2) treatment data from the Irish National Drug Treatment Reporting System (NDTRS) for 2015. The profiles of GPS cases with CUD and NDTRS cases were compared using 2-sided t tests designed for independent samples. Results: The prevalence of last year cannabis use among adults aged 15 and older was 6.5% and the prevalence of CUD was 2.6%, representing 94,515 of the Irish population. A total of 4,761 cases entered treatment for problem cannabis use. NDTRS treatment cases were significantly more likely than GPS cases to be unemployed (63.7% vs. 26.6%) and have no or primary level only educational attainment (56.3% vs. 21.2%). Over half (53.3%) of NDTRS cases first used cannabis before the age of 15 years, compared to 14.7% of CUD cases in the population. Discussion/Conclusion: Our findings suggest that earlier users and those with more complex or disadvantaged lives are more likely to seek treatment. A broad population health approach that engages multiple sectors such as health, social welfare, and education is recommended to ensure that there is increased opportunity for people with CUD to be identified and signposted towards treatment.

The determinants of debt financing: The case of fintech start-ups

The purpose of this paper is to explore determinants of the debt financing of FinTech start-ups. Using a new hand-collected multisource database that maps FinTech start-ups incorporated in the UK from 2010 to 2015, this study examines how the characteristics of FinTech start-ups affect the types of financing used in the first three years after incorporation. The novelty of this study consists in the identification and analysis of the determinants that enable FinTech start-ups to obtain long-term debt and hence to finance their growth. The analysis is primarily conducted via a Tobit regression model. This paper contributes to the literature since we still have limited understanding of the financing of FinTech firms, even if academic literature examining the financing of start-ups has expanded in the last few years. The results from the empirical analysis demonstrate that unregulated FinTech start-ups are more likely to be financed with long-term debt. Asset structure, owner characteristics and the specific FinTech activity influence the funding source. Moreover, FinTech start-ups backed by equity investors receive less long-term debt funding than their peers. A better understanding of the debt financing of FinTech start-ups provides managers with valuable insights into ways of managing their firms. Furthermore, our results have relevant implications for the employment, competition and innovation, given the role that start-ups generally play in the economy. The study is limited to a sample of FinTech start-ups incorporated in an advanced economy, and thus the generalization of the presented results to developing economies will require caution.

The Crisis of Social Institutions and Police Homicides: The Adverse Effects of Low Institutional Control

Using U.S. census data and a multi-source database on officer-involved killings, the current study extends previous research by exploring the influence of measures of weak social control in economic, educational, and familial institutions on state rates of police homicide. States with lower levels of institutional control are found to have higher overall rates of police homicides and police killings involving Black, Hispanic, and White decedents. The significant effects of institutional control on these police homicide rates are generally found to exhibit contextual invariance across different levels of various control variables (e.g., comparisons of states with low or high violent crime rates, low vs high economic inequality, low vs high levels of urbanization). These results and the limitations of this study are discussed in terms of implications for future research and public policy on police homicides and the role of social institutions in minimizing the occurrence of these incidents.

A Multimodality Information Synchronization Scheme for a Multisource Information System in the Electric Grid

A multisource and heterogeneous database is an important problem that disturbs the use of the electric power information system. The existing database synchronization scheme has some problems in practical applications, such as high resource loss and poor portability. This paper presents a high-efficiency database synchronization scheme for the electric power information system. The database is monitored, and its changes are captured by the shadow table and trigger method. Thus, data could be exchanged in trusted networks and nontrusted networks. In addition, a predetermined strategy is used to avoid data conflicts and ensure consistency and reliability of data synchronization. The above method is applied in the protection system of power networks. The results show that the synchronization scheme can effectively ensure the security of the system and has higher synchronization efficiency.

A Web Interface for Analyzing Hate Speech

Social media services make it possible for an increasing number of people to express their opinion publicly. In this context, large amounts of hateful comments are published daily. The PHARM project aims at monitoring and modeling hate speech against refugees and migrants in Greece, Italy, and Spain. In this direction, a web interface for the creation and the query of a multi-source database containing hate speech-related content is implemented and evaluated. The selected sources include Twitter, YouTube, and Facebook comments and posts, as well as comments and articles from a selected list of websites. The interface allows users to search in the existing database, scrape social media using keywords, annotate records through a dedicated platform and contribute new content to the database. Furthermore, the functionality for hate speech detection and sentiment analysis of texts is provided, making use of novel methods and machine learning models. The interface can be accessed online with a graphical user interface compatible with modern internet browsers. For the evaluation of the interface, a multifactor questionnaire was formulated, targeting to record the users’ opinions about the web interface and the corresponding functionality.

Is Europe facing an opioid crisis like the United States? An analysis of opioid use and related adverse effects in 19 European countries between 2010 and 2018.

Given the ongoing opioid crisis in the United States (US), we investigated the opioid situation in Europe. The aims of the study are to provide an overview of trends in prescription opioid (PO) use and opioid-related adversities between 2010 and 2018 for different opioids in 19 European countries and to present a comparison with similar data from the US. A multisource database study with national data from 19 European countries evaluating trends between 2010 and 2018 in (a) PO consumption, (b) high-risk (HR) opioid users, (c) opioid-related hospital admissions, (d) opioid-related overdose deaths, (e) opioid use disorder treatment entries, and (f) patients in opioid substitution therapy (OST). Within and between-country comparisons and comparisons with data from the US were made. There was considerable variation between European countries. Most countries showed increased PO consumption with the largest increase and the highest consumption in the United Kingdom (UK) compared to the rest of Europe and the US in 2018 (UK: 58,088 defined daily doses for statistical purposes/1000 population/day). In 2018, Scotland had the highest rates (per 100,000 population) of HR opioid users (16·2), opioid-related hospital admissions (118), opioid-related deaths (22·7), opioid use disorder treatment admissions (190), and OST patients (555) of all included European countries. These rates were similar or even higher than those in the US in 2018. Other countries with high rates of opioid-related adversities were Northern Ireland (synthetic and "other" opioids), Ireland (heroin and methadone), and England (all opioids). All other countries had no or little increase in opioid-related adversities. Apart from the British Isles and especially Scotland, there is no indication of an opioid crisis comparable to that in the US in the 19 European countries that were part of this study. More research is needed to identify drivers and develop interventions to stop the emerging opioid crisis in the UK and Ireland.

Ship Matching Using Convolutional Neural Network in Multi-source Synthetic Aperture Radar Images

Niu, L. and Lang, H., 2020. Ship matching using convolutional neural network in multi-source synthetic aperture radar images. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 166-175. Coconut Creek (Florida), ISSN 0749-0208.There are two major challenges in the ship matching and tracking task by synthetic aperture radar (SAR) images. First, images from different satellites have different incidence angles, polarization modes and resolutions, leading to the nonlinear geometric deformation of ships and the difference in image quality. The above differences result in a significant drop in the similarity of matching ships. Second, the principal direction of ships is arbitrary, which requires that the matching method has rotation-invariance. For the first challenge, the siamese-type convolution neural networks are adopted to reduce the distance between matching ships and increase the distance between non-matching ships. For the second challenge, a minimum bounding square (MBS) segmentation method is used to process all the ship images to the horizontal direction. After image preprocessing using MBS, there are only two situations for the principal direction of ships and rotation-invariance is transformed into the invariance of prow-to-stern interchange. To solve the problem of prow-to-stern interchange, a constrained hinge loss is proposed to ensure that the minimum similarity of matching images is still greater than the maximum similarity of non-matching images. There are a variety of siamese-type convolutional neural networks with good performance currently, the focus of the research is to use the two proposed strategies (MBS and constrained hinge loss) to improve the performance of existing methods instead of designing network architectures. In order to verify the feasibility of the proposed strategies, a multi-source ship target database covers different SAR sensors is constructed for ship matching. The experimental results prove that the proposed methods can match the ship target precisely and outperform the methods without MBS preprocessing and the constrained hinge loss function.