Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
128,616
result(s) for
"Digital data"
Sort by:
Convolutional Neural Network for Remote-Sensing Scene Classification: Transfer Learning Analysis
Remote-sensing image scene classification can provide significant value, ranging from forest fire monitoring to land-use and land-cover classification. Beginning with the first aerial photographs of the early 20th century to the satellite imagery of today, the amount of remote-sensing data has increased geometrically with a higher resolution. The need to analyze these modern digital data motivated research to accelerate remote-sensing image classification. Fortunately, great advances have been made by the computer vision community to classify natural images or photographs taken with an ordinary camera. Natural image datasets can range up to millions of samples and are, therefore, amenable to deep-learning techniques. Many fields of science, remote sensing included, were able to exploit the success of natural image classification by convolutional neural network models using a technique commonly called transfer learning. We provide a systematic review of transfer learning application for scene classification using different datasets and different deep-learning models. We evaluate how the specialization of convolutional neural network models affects the transfer learning process by splitting original models in different points. As expected, we find the choice of hyperparameters used to train the model has a significant influence on the final performance of the models. Curiously, we find transfer learning from models trained on larger, more generic natural images datasets outperformed transfer learning from models trained directly on smaller remotely sensed datasets. Nonetheless, results show that transfer learning provides a powerful tool for remote-sensing scene classification.
Journal Article
Species interactions: next‐level citizen science
We envisage a future research environment where digital data on species interactions are easily accessible and comprehensively cover all species, life stages and habitats. To achieve this goal, we need data from many sources, including the largely untapped potential of citizen science for mobilising and utilising existing information on species interactions. Traditionally volunteers contributing information on the occurrence of species have focused on single‐species observations from within one target taxon. We make recommendations on how to improve the gathering of species interaction data through citizen science, which data should be collected and how it can be motivated. These recommendations include providing feedback in the form of network visualisations, leveraging a wide variety of other data sources and eliciting an emotional connection to the species in question. There are many uses for these data, but in the context of biological invasions, information on species interactions will increase understanding of the effects of invasive alien species on recipient communities and ecosystems. We believe that the inclusion of ecological networks as a concept within citizen science, not only for initiatives focussed on biological invasions but also across other ecological themes, will not only enrich scientific knowledge on species interactions but also deepen the experience and enjoyment of citizens themselves.
Journal Article
Studying illicit drug trafficking on Darknet markets: Structure and organisation from a Canadian perspective
•Study of illicit drug trafficking using data collected on 8 cryptomarkets.•Knowledge on the structure and organisation of distribution networks.•Use of an approach combining the analysis of vendor names and PGP keys.•Results reveal the presence of key actors of the Canadian illicit drug trafficking.
Cryptomarkets are online marketplaces that are part of the Dark Web and mainly devoted to the sale of illicit drugs. They combine tools to ensure anonymity of participants with the delivery of products by mail to enable the development of illicit drug trafficking.
Using data collected on eight cryptomarkets, this study provides an overview of the Canadian illicit drug market. It seeks to inform about the most prevalent illicit drugs vendors offer for sale and preferred destination countries. Moreover, the research gives an insight into the structure and organisation of distribution networks existing online. In particular, we provide information about how vendors are diversifying and replicating across marketplaces. We inform on the number of listings each vendor manages, the number of cryptomarkets they are active on and the products they offer.
This research demonstrates the importance of online marketplaces in the context of illicit drug trafficking. It shows how the analysis of data available online may elicit knowledge on criminal activities. Such knowledge is mandatory to design efficient policy for monitoring or repressive purposes against anonymous marketplaces. Nevertheless, trafficking on Dark Net markets is difficult to analyse based only on digital data. A more holistic approach for investigating this crime problem should be developed. This should rely on a combined use and interpretation of digital and physical data within a single collaborative intelligence model.
Journal Article
Digital transformation risk management in forensic science laboratories
•Examples of digital transformation difficulties encountered in laboratory environments.•Robust risk mitigation strategies for managing digital transformations in forensic laboratories.•Forensic digital preparedness applied to forensic laboratories.•Role of digital forensic expertise in risk management of digital transformations in laboratories.•Recommended enhances to international quality standards such as ISO/IEC 17025.
Technological advances are changing how forensic laboratories operate in all forensic disciplines, not only digital. Computers support workflow management, enable evidence analysis (physical and digital), and new technology enables previously unavailable forensic capabilities. Used properly, the integration of digital systems supports greater efficiency and reproducibility, and drives digital transformation of forensic laboratories. However, without the necessary preparations, these digital transformations can undermine the core principles and processes of forensic laboratories. Pertinent examples of problems involving technology that have occurred in laboratories are provided, along with opportunities and risk mitigation strategies, based on the authors’ experiences. Forensic preparedness concentrating on digital data reduces the cost and operational disruption of responding to various kinds of problems, including misplaced exhibits, allegations of employee misconduct, disclosure requirements, and information security breaches. This work presents recommendations to help forensic laboratories prepare for and manage these risks, to use technology effectively, and ultimately strengthen forensic science. The importance of involving digital forensic expertise in risk management of digital transformations in laboratories is emphasized. Forensic laboratories that do not adopt forensic digital preparedness will produce results based on digital data and processes that cannot be verified independently, leaving them vulnerable to challenge. The recommendations in this work could enhance international standards such as ISO/IEC 17025 used to assess and accredit laboratories.
Journal Article
Social media forensics applied to assessment of post–critical incident social reaction: The case of the 2017 Manchester Arena terrorist attack
•Tracking different types of reactions on social media after a critical incident.•LDA topic modeling techniques are applied to 4M+ tweets corpus.•Tweets gave insights on the context, including requests for help and warnings.•A machine learning method that can enhance forensic investigations on digital traces.
Forensic science is constantly evolving and transforming, reflecting the numerous technological innovations of recent decades. There are, however, continuing issues with the use of digital data, such as the difficulty of handling large-scale collections of text data. As one way of dealing with this problem, we used machine-learning techniques, particularly natural language processing and Latent Dirichlet Allocation (LDA) topic modeling, to create an unsupervised text reduction method that was then used to study social reactions in the aftermath of the 2017 Manchester Arena bombing. Our database was a set of millions of messages posted on Twitter in the first 24 h after the attack. The findings show that our method improves on the tools presently used by law enforcement and other agencies to monitor social media, particularly following an event that is likely to create widespread social reaction. For example, it makes it possible to track different types of social reactions over time and to identify subevents that have a significant impact on public perceptions.
Journal Article