Data-intensive Methodology Research Articles

Data-intensive sociolinguistics using social media

This article looks into using large-scale social media data in SSH research and in particular in studies of language variation and change. It presents a case study that investigates the role of social networks in linguistic variability. Previous studies have convincingly shown that networks in which people are connected to each other in loose ties tend to contribute positively to language change. Conversely, networks in which people are closely connected to each other inhibit change. This conclusion is, however, based on small datasets, and this study tests if the difference is diluted when network size is closer to human average. The results suggest this to be the case. Towards the end, the article suggests numerous ways in which large-scale social media data and the use of data intensive methodologies could be increased and encouraged in SSH research.

UK health researchers’ considerations of the environmental impacts of their data-intensive practices and its relevance to health inequities

BackgroundThe health sector aims to improve health outcomes and access to healthcare. At the same time, the sector relies on unsustainable environmental practices that are increasingly recognised to be catastrophic threats to human health and health inequities. As such, a moral imperative exists for the sector to address these practices. While strides are currently underway to mitigate the environmental impacts of healthcare, less is known about how health researchers are addressing these issues, if at all.MethodsThis paper uses an interview methodology to explore the attitudes of UK health researchers using data-intensive methodologies about the adverse environmental impacts of their practices, and how they view the importance of these considerations within wider health goals.ResultsInterviews with 26 researchers showed that participants wanted to address the environmental and related health harms associated with their research and they reflected on how they could do so in alignment with their own research goals. However, when tensions emerged, their own research was prioritised. This was related to their own desires as researchers and driven by the broader socio-political context of their research endeavours.ConclusionTo help mitigate the environmental and health harms associated with data-intensive health research, the socio-political context of research culture must be addressed.

Use of nested multi-regional input–output analysis for the evaluation of subnational emission reduction policies: A case of Japanese prefectures

Many subnational regions around the world, in their own capacity for policy formulation and implementation, have set targets for net-zero carbon dioxide or greenhouse gas emissions. However, most targets do not consider consumption-based emissions, which include emissions generated outside their borders to produce the products they consume. It is partly due to methodological difficulties and data gaps that subnational governments often face but this situation could misguide their efforts in the face of the goal of reducing global emissions. To facilitate an understanding of consumption-based subnational carbon emissions and informed subnational policymaking, this study proposes a relatively simple and less data-intensive methodology for subnational consumption-based emission analysis. It uses a nested multi-regional input–output approach by combining a publicly available global multi-regional input–output table and a subnational input–output table, whose computation is repeated for multiple regions for comparison. A case study of three Japanese prefectures with different levels of renewable electricity consumption was conducted using the 2015 datasets. The results showed that two prefectures with similar socioeconomic conditions, except for their totally different renewable electricity ratios (90 % and 4.6 %), exhibited a 25 % difference (4.4- and 6.0-ton CO2) in per capita household consumption-based carbon dioxide emissions in 2015. Counterfactual policy scenarios were analyzed using the constructed models. Application of an increased renewable electricity scenario indicated that an approximate 10 % increase in the prefectural renewable electricity ratio could reduce household per capita consumption-based emissions by 151–179 kg CO2 in those prefectures. Consumption-based emissions disaggregated by product category indicated that increasing the renewable electricity ratio had significant consumption-based emission reduction effects on not only electricity itself but also on services, particularly commerce, hotel, and restaurant services. The potential impacts of changing renewable electricity ratios on consumption-based emissions were found to be high for products whose share in household spending on local products and electricity cost in their production was large. A scenario analysis of home electrification, assuming the total replacement of household fuel gases and kerosene with electricity, resulted in a modest consumption-based emission reduction potential of less than 10 % for a prefecture with a high renewable electricity ratio and even showed a slight emission increase for a prefecture with a low renewable electricity ratio. In using commonly available datasets without the need to model trade flows between subnational regions, this study's method could realistically be applied by subnational governments to estimate consumption-based emissions of their jurisdiction, compared with other regions, and analyze the potential effects of emission reduction policy scenarios.

Development and application of a new mangrove vegetation index (MVI) for rapid and accurate mangrove mapping

Advancement in Remote Sensing allows rapid mangrove mapping without the need for data-intensive methodologies, complex classifiers, and skill-dependent classification techniques. This study proposes a new index, the Mangrove Vegetation Index (MVI), to rapidly and accurately map mangroves extent from remotely-sensed imageries. The MVI utilizes three Sentinel-2 bands green, Near Infrared (NIR) and Shortwave Infrared (SWIR) in the form |NIR-Green|/|SWIR-Green| to discriminate the distinct greenness and moisture of mangroves from terrestrial vegetation and other land cover. Spectral band analysis shows that the |NIR-Green| part of MVI captures the differences of greenness between mangrove forests and terrestrial vegetation. The |SWIR-Green| part of the index expresses the distinct moisture of mangroves without the need for additional intertidal data and water indices. The MVI value increases with the increasing probability of a pixel being classified as mangroves. Eleven mangrove forests in the Philippines and one mangrove park in Japan were then mapped using MVI. Accuracy assessment was done using field inventory data and high-resolution drone orthophotos. MVI have successfully separated the mangroves from other cover especially terrestrial vegetation, with an overall index accuracy of 92%. The MVI was applied to Landsat 8 images using the equivalent bands to test the universality of the index. Comparable MVI mangrove maps were produced between Sentinel-2 and Landsat images, with an optimal minimum threshold of 4.5 and 4.6, respectively. MVI can effectively highlight the greenness and moisture information in mangroves as reflected by its moderate to high correlation value (r = 0.63 and 0.84, α = 0.05) with the Sentinel-derived chlorophyll-a (Ca) and canopy water (Cw) biophysical products. This study developed and implemented two automated platforms: an offline IDL-based ‘MVI Mapper’ and an online Google Earth Engine-based MVI mapping interface. The MVI implemented in Google Earth Engine was used in generating the latest mangrove extent map of the Philippines. Additionally, the application of MVI were tested to four additional mangrove forests in Southeast Asia: Thailand, Vietnam, Indonesia and Cambodia; and to selected mangroves forests in South America, Africa and Australia.

Siting of Rainwater Harvesting Potential Sites in Arid or Semi-arid Watersheds Using GIS-based Techniques

Rainwater harvesting (RWH) has emerged as one of the efficient strategies to augment surface and groundwater sources to meet water resource demands. The aim of this study is to present a methodology for siting RWH areas and selecting appropriate RWH structures in a watershed using Geographic Information Systems (GIS) and Remote Sensing (RS) based multi-criteria analysis. Derived thematic layers such as drainage density, slope, road network, runoff potential, distance to streams, and distance to built-up areas were used to identify suitable areas for RWH. The Weighted Linear Combination (WLC) technique was used in combination with the Boolean approach in a GIS environment. Normalized weights assigned to thematic layers and features were tested for consistency using the consistency ratio in Analytical Hierarchy Process. The study area can be divided into three categories of suitability, i.e.: (i) Most suitable with 26% coverage (167 km2); (ii) moderately suitable with 30% coverage; and (iii) least suitable with 12.6% coverage (80 km2) of the study area. Suitable RWH and artificial recharge structures were identified based on identified zones. It was found that 11 check dams could be implemented in the study area and 3 percolation tanks along streams and on the ground. This cost-effective and low data-intensive methodology presented can be replicated by relevant stakeholders in arid regions to adopt RWH as an effective strategy to address growing water shortages.

Data Tagging in Medical Images: A Survey of the State-of-Art.

A huge amount of medical data is generated every second, and a significant percentage of the data are images that need to be analyzed and processed. One of the key challenges in this regard is the recovery of the data of medical images. The medical image recovery procedure should be done automatically by the computers that are the method of identifying object concepts and assigning homologous tags to them. To discover the hidden concepts in the medical images, the lowlevel characteristics should be used to achieve high-level concepts and that is a challenging task. In any specific case, it requires human involvement to determine the significance of the image. To allow machine-based reasoning on the medical evidence collected, the data must be accompanied by additional interpretive semantics; a change from a pure data-intensive methodology to a model of evidence rich in semantics. In this state-of-art, data tagging methods related to medical images are surveyed which is an important aspect for the recognition of a huge number of medical images. Different types of tags related to the medical image, prerequisites of medical data tagging, different techniques to develop medical image tags, different medical image tagging algorithms and different tools that are used to create the tags are discussed in this paper. The aim of this state-of-art paper is to produce a summary and a set of guidelines for using the tags for the identification of medical images and to identify the challenges and future research directions of tagging medical images.

The DARA Big Data Project

AbstractThe DARA Big Data project is a flagship UK Newton Fund & GCRF program in partnership with the South African Department of Science & Technology (DST). DARA Big Data provides bursaries for students from the partner countries of the African VLBI Network (AVN), namely Botswana, Ghana, Kenya, Madagascar, Mauritius, Mozambique, Namibia and Zambia, to study for MSc(R) and PhD degrees at universities in South Africa and the UK. These degrees are in the three data intensive DARA Big Data focus areas of astrophysics, health data and sustainable agriculture. The project also provides training courses in machine learning, big data techniques and data intensive methodologies as part of the Big Data Africa initiative.

Investigating transparency regarding ecoinvent users’ system model choices

Life cycle assessment (LCA) is a data-intensive methodology; therefore, experts usually focus collection efforts on a few activities, while generic data on remaining activities are taken from databases. Even though increased availability of databases has facilitated LCA takeoff, assuring data quality is fundamental to ensure meaningful results and reliable interpretation. Ecoinvent has become a global reference for inventory data. Its current version released three impact partition modeling options—the recycled content, “allocation at the point of substitution” (APOS), and consequential models—whose adequate choice is crucial for yielding meaningful assessments. Tutorials and manuals describe the distribution algorithm that backs each system model, to ground decision-making regarding the best fit to a study’s goals. We performed a systematic literature review to investigate—within the papers published on the International Journal of LCA (IJLCA)—how transparently authors addressed the system model choices. About 70% of LCA practitioners continued to use earlier versions of ecoinvent after version 3 was launched in 2013. The number of papers using versions 3.x only showed an increased growth trend 2 years later. Eighty-three papers actually adopted the newest version of the database. From those, only 29 papers clearly mentioned the adopted system model. Our SLR also suggests a trend regarding authorship profile of LCA-related studies: the number of studies conducted by practitioners aware of the intricacies of sound modeling of background and foreground data might have been surpassed by those conducted by non-LCA specialists who use LCA as a supporting tool for investigations in applied fields, and merely scratch the surface. Our results point to a need for a caveat: ecoinvent users must take time to understand the general concept behind each system model and practice one of the most important actions when performing an LCA—state methodological choices clearly.

Energy rating of a water pumping station using multivariate analysis

Among water management policies, the preservation and the saving of energy demand in water supply and treatment systems play key roles. When focusing on energy, the customary metric to determine the performance of water supply systems is linked to the definition of component-based energy indicators. This approach is unfit to account for interactions occurring among system elements or between the system and its environment. On the other hand, the development of information technology has led to the availability of increasing large amount of data, typically gathered from distributed sensor networks in so-called smart grids. In this context, data intensive methodologies address the possibility of using complex network modeling approaches, and advocate the issues related to the interpretation and analysis of large amount of data produced by smart sensor networks.In this perspective, the present work aims to use data intensive techniques in the energy analysis of a water management network. The purpose is to provide new metrics for the energy rating of the system and to be able to provide insights into the dynamics of its operations. The study applies neural network as a tool to predict energy demand, when using flowrate and vibration data as predictor variables.

The Paradigm Shift to Multimodality in Contemporary Computer Interfaces

During the last decade, cell phones with multimodal interfaces based on combined new media have become the dominant computer interface worldwide. Multimodal interfaces support mobility and expand the expressive power of human input to computers. They have shifted the fulcrum of human-computer interaction much closer to the human. This book explains the foundation of human-centered multimodal interaction and interface design, based on the cognitive and neurosciences, as well as the major benefits of multimodal interfaces for human cognition and performance. It describes the data-intensive methodologies used to envision, prototype, and evaluate new multimodal interfaces. From a system development viewpoint, this book outlines major approaches for multimodal signal processing, fusion, architectures, and techniques for robustly interpreting users' meaning. Multimodal interfaces have been commercialized extensively for field and mobile applications during the last decade. Research also is growing rapidly in areas like multimodal data analytics, affect recognition, accessible interfaces, embedded and robotic interfaces, machine learning and new hybrid processing approaches, and similar topics. The expansion of multimodal interfaces is part of the long-term evolution of more expressively powerful input to computers, a trend that will substantially improve support for human cognition and performance.

Data-intensive modeling of forest dynamics

Forest dynamics are highly dimensional phenomena that are not fully understood theoretically. Forest inventory datasets offer unprecedented opportunities to model these dynamics, but they are analytically challenging due to high dimensionality and sampling irregularities across years. We develop a data-intensive methodology for predicting forest stand dynamics using such datasets. Our methodology involves the following steps: 1) computing stand level characteristics from individual tree measurements, 2) reducing the characteristic dimensionality through analyses of their correlations, 3) parameterizing transition matrices for each uncorrelated dimension using Gibbs sampling, and 4) deriving predictions of forest developments at different timescales. Applying our methodology to a forest inventory database from Quebec, Canada, we discovered that four uncorrelated dimensions were required to describe the stand structure: the biomass, biodiversity, shade tolerance index and stand age. We were able to successfully estimate transition matrices for each of these dimensions. The model predicted substantial short-term increases in biomass and longer-term increases in the average age of trees, biodiversity, and shade intolerant species. Using highly dimensional and irregularly sampled forest inventory data, our original data-intensive methodology provides both descriptions of the short-term dynamics as well as predictions of forest development on a longer timescale. This method can be applied in other contexts such as conservation and silviculture, and can be delivered as an efficient tool for sustainable forest management.

Understanding the population dynamics of a rare, polyvoltine butterfly

AbstractUnderstanding the effects of endogenous and exogenous factors on population dynamics is essential for assessing the viability of populations, setting recovery goals for endangered species, and evaluating management options. Invertebrates are particularly difficult to monitor and few long‐term datasets are available for these species. Additionally, limited resources make it necessary to perform monitoring as efficiently as possible. Here, I use the bivoltine Karner blue butterfly Lycaeides melissa samuelis to demonstrate how analyzing the effects of density‐dependent factors and weather on separate life stages can be utilized to understand monitoring data, assess populations and identify critical life‐history parameters. My first step was to compare the use of peak numbers as an index of population size with estimates obtained from a more data intensive methodology. Peak numbers proved to be an effective index, and so I utilized this index to analyze 10 and 13 years of monitoring data at two Karner blue butterfly sites in New York, USA. I modeled the effects of weather and density dependence on two distinct population growth rates (λ) per year. Analysis with Akaike's Information Criteria indicated that both sites were primarily influenced by density dependence during the summer period and by weather conditions during the winter period. Large population declines occurred in the winter period and were a result of the previous year's dry summer and cool spring weather. I conclude that recovery goals for this endangered species should include a second brood‐carrying capacity, mean winter growth rate and multiple sites with independent populations. This study represents a rare, long‐term study on the population dynamics of a polyvoltine species. Understanding the population dynamics of polyvoltine species, such as the Karner blue butterfly, will assist in the conservation of many invertebrate and small mammal species.