Online Domain Research Articles

Genealogies of online content identification - an introduction

In today’s digital landscape, cultural content such as texts, films, images, and recorded sounds are increasingly subjected to automatic (or semi-automatic) processes of identification and classification. On a daily basis, spam filters scan swaths of emails in order to separate legit and illegitimate textual messages (Brunton, 2013), algorithms analyze years of user-uploaded film on YouTube in search for copyright violations (Heuguet, 2019), and software systems are deployed to scrutinize millions of images on social media sites in order to detect sexually offensive content (Liao, 2018). These examples reveal how machines and algorithmic systems are increasingly utilized to make complex judgments regarding cultural content. Indeed, it could be argued that the wide-ranging adoption of content identification systems is constructing new ontologies of culture and regimes of truth in the online domain. When put to action, content identification systems are trusted with the ability to separate good/bad and legal/illegal forms of communication and used to secure the singularity, value, authenticity, origin, and ownership of content. Such efforts are deeply embedded in constructions of knowledge, new forms of political governance, and not least global market transactions. Content identification tools now make up an essential part of the online data economy by protecting the interests of rights holders and forwarding the mathematization, objectification, and commodification of cultural productions.

The online behaviors of Islamic state terrorists in the United States

Research SummaryThis study offers an empirical insight into terrorists’ use of the Internet. Although criminology has previously been quiet on this topic, behavior‐based studies can aid in understanding the interactions between terrorists and their environments. Using a database of 231 US‐based Islamic State terrorists, four important findings are offered: (1) This cohort utilized the Internet heavily for the purposes of both networking with co‐ideologues and learning about their intended activity. (2) There is little reason to believe that these online interactions are replacing offline ones, as has previously been suggested. Rather, terrorists tend to operate in both domains. (3) Online activity seems to be similar across the sample, regardless of the number of co‐offenders or the sophistication of attack. (4) There is reason to believe that using the Internet may be an impediment to terrorists’ success.Policy ImplicationsThe findings of this study have two important policy implications. First, it is vital to understand the multiplicity of environments in which terrorists inhabit. Policy makers have tended to emphasize the online domain as particularly dangerous and ripe for exploitation. While this is understandable from one perspective, simplistic and monocausal explanations for radicalization must be avoided. Terrorists operate in both the online and offline domain and there is little reason to believe that the former is replacing the latter. The two may offer different criminogenic inducements to would‐be terrorists, and at times they may be inseparably intertwined. Second, when policy responses do focus on online interventions, it is vital to understand the unintended consequences. This is particularly the case for content removal, which may inadvertently be aiding terrorists and hampering law enforcement investigations.

The use of Virtual Media in Postgraduate Medical Training During the COVID-19 Pandemic

Context: The novel coronavirus disease 2019 (COVID-19) pandemic brought about public health safety measures of social distancing and avoidance of crowded gatherings. These led to widespread disruptions in activities of daily living, as well as routine health-care services and learning activities. There has been a noticeable move to the online domain as the platform for teaching and learning. Aims: The present study aimed to describe the online learning and teaching experiences of postgraduate medical doctors in Nigerian hospitals. Subjects and Methods: The study was a cross-sectional online survey sent to medical doctors across social media platforms, and it sought to find out the current number of hours utilized and modalities of their online educational media use. Results: A total of 116 doctors responded and completed the survey, with 74 (63.8%) and 104 (89.6%) of them having some webinar experience before and during the pandemic, respectively. At least 82 (70.4%) had 1–5 h of online learning activities each week, and an overwhelming 114 (98.2%) used a personal Internet data subscription to access online educational material. In addition, 56 (48.3%) felt that online learning was less rewarding than in-person interactions even though 93 (80.2%) found the online learning activities to be quite convenient. Poor Internet connectivity was cited as a major challenge to a successful online learning experience in 80 (68.9%). Conclusion: Although the COVID-19 pandemic restricted face-to-face human contact including learning activities, postgraduate medical education has filled the void by increasing online learning activities, resulting in an admixture of experiences and reactions.

Incremental Real-Time Personalization in Human Activity Recognition Using Domain Adaptive Batch Normalization

Human Activity Recognition (HAR) from devices like smartphone accelerometers is a fundamental problem in ubiquitous computing. Machine learning based recognition models often perform poorly when applied to new users that were not part of the training data. Previous work has addressed this challenge by personalizing general recognition models to the unique motion pattern of a new user in a static batch setting. They require target user data to be available upfront. The more challenging online setting has received less attention. No samples from the target user are available in advance, but they arrive sequentially. Additionally, the motion pattern of users may change over time. Thus, adapting to new and forgetting old information must be traded off. Finally, the target user should not have to do any work to use the recognition system by, say, labeling any activities. Our work addresses all of these challenges by proposing an unsupervised online domain adaptation algorithm. Both classification and personalization happen continuously and incrementally in real time. Our solution works by aligning the feature distributions of all subjects, be they sources or the target, in hidden neural network layers. To this end, we normalize the input of a layer with user-specific mean and variance statistics. During training, these statistics are computed over user-specific batches. In the online phase, they are estimated incrementally for any new target user.

The media epistemic value of sonic analytics tools. A commentary

While automatized content identification of audio data, in critical discourse analysis, is bound to the symbolic order of monitoring, control, surveillance, censorship and copyright protection, the very tools and algorithms which have been developed for such purposes can be turned into instruments of knowledge production in the scientific sense. Audio content identification is not simply an extension of cultural taxonomies to machine listening, but an operation with its own eigen knowledge. Audio content identification is not simply a continuation of analog techniques for monitoring sonic objects. From a media-epistemological perspective, new forms of audio content identification open different orders of the sonic archive. What is practiced in the online domain has been preceded by experimental investigations of archival storage. The real l'archive, though, are the technological (hardware) and mathematical (software) criteria defining content identification. A media archaeology of audio content identification reveals the technological l'archive governing such forms of enunciation.

The Fuzzy Experiment Approach for Detection and Prevention of Phishing attacks in online Domain

The Fuzzy Experiment Approach for Detection and Prevention of Phishing attacks in online Domain

Night-to-Day: Online Image-to-Image Translation for Object Detection Within Autonomous Driving by Night

Object detectors are central to autonomous driving and are widely used in driver assistance systems. Object detectors are trained on a finite amount of data within a specific domain, hampering detection performance when applying object detectors to samples from other domains during inference, an effect known as domain gap. Domain gap is a concern for data-driven applications, evoking repetitive retraining of networks when the applications unfold into other domains. With object detectors that have been trained on day images only, a domain gap can be observed in object detection by night. Training object detectors on night images is critical because of the enormous effort required to generate an adequate amount of diversely labeled data, and existing data sets often tend to overfit specific domain characteristics. For the first time, this work proposes adapting domains by online image-to-image translation to expand an object detector's domain of operation. The domain gap is decreased without additional labeling effort and without having to retrain the object detector while unfolding into the target domain. The approach follows the concept of domain adaptation, shifting the target domain samples into the domain knownto the object detector (source domain). Firstly, the UNIT network is trained for domain adaptation and subsequently cast into an online domain adaptation module, which narrows down the domain gap. Domain adaptation capabilities are evaluated qualitatively by displaying translated samples and visualizing the domain shift through the 2D tSNE algorithm. We quantitatively benchmark the domain adaptation's influence on a state-of-the-art object detector, and on a retrained object detector, for mean average precision, mean recall, and the resulting F1-score. Our approach achieves an F1 score improvement of 5.27 % within object detection by night when applying online domain adaptation. The evaluation is executed on the BDD100K benchmark data set.

An Investigation of What Factors Determine the Way in Which Customer Satisfaction Is Increased through Omni-Channel Marketing in Retail

There have been significant changes evident across the retail sector over the last 20 years, which have been primarily influenced by technological advancements. These changes have forced many retailers to rethink their business models and utilise a combination of both brick and mortar stores and online domains to present an integrated shopping experience. In order to gain an understanding of how customer satisfaction is increased through omni-channel retailing, we collected data through survey research, specifically through semi-structured interviews and questionnaires. In order to generate concise and accurate findings, the data were analysed using thematic analysis. From this analysis, three themes were generated. The first theme explores the reasoning behind preferred shopping methods and what influences these customers’ purchase methods. The second theme explores customers’ personal preference and investigates the potential of using virtual reality as a method for increasing customer satisfaction. Finally, the third theme looks into utilising interactive technology to create a diversified user experience. This theme explores the potential of utilising a communication platform for customers and also designing a tangible interface for disabled customers. This study has opened up an opportunity for future research and has outlined potential design insights in how to improve customer experience and customer purchase intentions.

Development of Web-based Technology for Environmental Engineering Education System

The objective of the ecological building instruction ought not just train understudies' natural information, the more significant thing is that it prepares understudies' natural ethics and structures the conduct which is good for the earth, and these must be shaped by training, in actuality. In the customary showing model of training, one instructor can just guide one practice simultaneously. With the improvement of organization innovation, instructor can control the distinctive practice exercises firing up in various areas or in various occasions by network. In light of the incorporation of viable need and intuitive qualities of condition instruction, the creator set forward an online domain training mode named "practice-intelligent partake in". The Core of this mode is to prepare understudies' natural ethics by training and to understand educators' guidance through organization.

Bi-directional online transfer learning: a framework

Transfer learning uses knowledge learnt in source domains to aid predictions in a target domain. When source and target domains are online, they are susceptible to concept drift, which may alter the mapping of knowledge between them. Drifts in online environments can make additional information available in each domain, necessitating continuing knowledge transfer both from source to target and vice versa. To address this, we introduce the Bi-directional Online Transfer Learning (BOTL) framework, which uses knowledge learnt in each online domain to aid predictions in others. We introduce two variants of BOTL that incorporate model culling to minimise negative transfer in frameworks with high volumes of model transfer. We consider the theoretical loss of BOTL, which indicates that BOTL achieves a loss no worse than the underlying concept drift detection algorithm. We evaluate BOTL using two existing concept drift detection algorithms: RePro and ADWIN. Additionally, we present a concept drift detection algorithm, Adaptive Windowing with Proactive drift detection (AWPro), which reduces the computation and communication demands of BOTL. Empirical results are presented using two data stream generators: the drifting hyperplane emulator and the smart home heating simulator, and real-world data predicting Time To Collision (TTC) from vehicle telemetry. The evaluation shows BOTL and its variants outperform the concept drift detection strategies and the existing state-of-the-art online transfer learning technique.

Internet since decades - but no globally searchable data

The usability of digital information on the internet strongly depends on its searchability. There are search engines for global language based text search, but language based information representation has only limited value, e.g. concerning reproducibility, comparability, precision. Searchable digital information can be much more. Any piece of digital information is a number sequence. We can define the binary format and value set resp. domain of every number online. Then universal information transport is possible by the online defined "Domain Vector" (DV) data structure: "UL plus number sequence". At this "UL" is an efficient link to the machine readable online definition of the number sequence. The UL also is global identifier of a certain kind of data. The online definition includes additional information, e.g. for similarity comparison of the number sequence. With this a universal numeric search engine can provide precise user defined worldwide search of DVs like in a globalized database. All users can participate. Together we can optimize online definitions and provide defined data as DVs. Initially it is necessary to build a first attractive online presence where users can provide online definitions. We should stay in contact to optimize together the common standard for machine readable online definitions and DVs.

Designing an Integrated Multi-Channel Multi Product Closed Loop Sustainable Supply Chain Network Meeting Customer Requirements

Indian manufacturing firms which started by focusing on offline exclusive sales (single channel), are expanding their sales to online domain (multichannel) today. They are still facing a lot of difficulties in interfacing various channels efficiently in their multichannel supply chain network, this integration of different channels increases its customers range to purchase and customer comfort. This paper aims to develop a multiproduct closed loop supply chain model for an Indian kitchenware firm to address their integrating issues between multiple channels take into consideration the environmental and economic sustainability. IBM CPLEX optimiser is used to solve the model for minimising the transportation cost between various stages of supply chain. To take into account the real world factors of customer behaviour, four types of customer zones are selected to accommodate different customer choices. The results show how the optimal supply chain structure is affected by these multichannel customer preferences.

COVID-19 and digital disruption in UK universities: afflictions and affordances of emergency online migration

COVID-19 has caused the closure of university campuses around the world and migration of all learning, teaching, and assessment into online domains. The impacts of this on the academic community as frontline providers of higher education are profound. In this article, we report the findings from a survey of n = 1148 academics working in universities in the United Kingdom (UK) and representing all the major disciplines and career hierarchy. Respondents report an abundance of what we call ‘afflictions’ exacted upon their role as educators and in far fewer yet no less visible ways ‘affordances’ derived from their rapid transition to online provision and early ‘entry-level’ use of digital pedagogies. Overall, they suggest that online migration is engendering significant dysfunctionality and disturbance to their pedagogical roles and their personal lives. They also signpost online migration as a major challenge for student recruitment, market sustainability, an academic labour-market, and local economies.

Self-reported motivations for engaging or declining to engage in cyber-dependent offending and the role of autistic traits

Cyber-dependent offending, i.e. criminal behaviour reliant on computing and the online domain, has been reportedly associated with particular characteristics and motivations such as being young, male, autistic and motivated by challenge. These associations are anecdotal however and empirical evidence is limited. The present study investigated reasons for engaging or declining to commit cyber-dependent offending in cyber-skilled non-offenders (n = 175) and offenders (n = 7) via an online survey measuring cyber-dependent criminality. The potential role of autism and autistic traits was also considered. Qualitative interviews about motivations for offending were carried out with the offenders. The cyber-dependent offenders reported seven main reasons for engaging in cyber-dependent offending: (1) lack of understanding; (2) entertainment; (3) peer influence; (4) experience and career; (5) anonymity and risk perception; (6) life events; and (7) morals. Twenty-nine (approximately 17 %) of the non-offenders had been asked to engage in cyber-dependent offending but had declined. Their reasons and motivations for declining to commit cyber-dependent offences were compared with the cyber-dependent offenders reasons and motivations for engaging in cybercrime. Seven main reasons for declining to offend were identified: (1) moral principles; (2) perception of risk; (3) fear of consequences; (4) not wanting to; (5) wanting to adhere to the law; (6) behaviour being too complicated; and (7) price being too low. Implications for practise are discussed.

Trolling as a Collective Form of Harassment: An Inductive Study of How Online Users Understand Trolling

Rising media and academic concerns for the social implications of online trolling require that scholars understand how everyday users conceptualize trolling. The validity of survey instruments may potentially be at risk if online users and academics, who subsequently advocate for interventions, have competing understandings of trolling. Surveying 120 US-based online users, I find that the spectrum of behaviors classified in the literature as “trolling” do not resonate with respondents. Instead, respondents report that harassment on the basis of race and gender is indicative of trolling. These results suggest a disconnect between academic definitions of trolling that focus on prosocial or humor-based forms of trolling and lay definitions which foreground identity-based harassment and harm. Current instruments that assess trolling behaviors and experiences do not attend to this definition of trolling as identity-based harassment, calling into question construct validity. Furthermore, sociologists identify identity-based harassment as a form of discrimination in other domains of social life, but present measures of discrimination do not consider the online domain. This suggests that studies of everyday discrimination and the implications of mistreatment should extend to online spaces, where racism and sexism are reproduced.

Two-phase methodology for prioritization and utility assessment of software vulnerabilities

There seems to be a system or piece of software for everything nowadays—from an application that lets you explore internet browsers to virtual reality software. This growth in the online domain is pushing the software developers to create safer products and enhance the protection of sensitive user information. This security issue is becoming even more critical due to the rise of cyber-attacks; the industry hence is giving foremost attention to the testing and development phase. Since there can be several flaws or weaknesses in the software or an operating system, that could allow an attacker to compromise the integrity, availability, or confidentiality of a product, commonly termed as software vulnerabilities. Thus, the first step is to understand which type of vulnerability is the most crucial to minimize the losses. Multi-criteria decision-making techniques, therefore, help to prioritize these vulnerabilities. In this paper, we apply two-phase methodology, comprising of the analytic hierarchy process (AHP) and best–worst method (BWM) in the first phase followed by the two-way assessment technique in the second phase, which helps in assessing the vulnerabilities in terms of their utility. The model is validated using real-life data of a software testing and development company situated in the northern part of India. The results show that the BWM performed significantly better as compared to the AHP approach in terms of utility.

Gender Differences in Social Support on Social Network Sites: A Meta-Analysis.

Social network sites (SNS) have become an increasingly popular platform for providing and receiving social support. Traditionally, females are believed to offer more social support. However, investigations of both the offline and online domains have revealed mixed findings. This meta-analysis attempts to assess the presence and magnitude of gender differences in social support on SNS. A literature review produced a sample of 30 independent studies with 17,000 participants. Results showed that females on SNS give (d = 0.36) and receive (d = 0.14) greater social support than do males. This is the first meta-analysis to test for and demonstrate gender differences in social support, either offline or online.

Monetisation of Fake News in the Cyber Domain: A Roadmap for Building Domestic and International Cyber Resilience

The online domain has witnessed the proliferation of fake news in the past few years all over the world. This overwhelming amount of (dis)information has created incentives for the proliferators of fake news to benefit from ongoing political developments across the world, due to the courtesy of the existing model of online revenue generation on the Internet. The unregulated, liberal and open nature of the Internet has created an unintended consequence of incentive maximisation for fake news. This paper takes an analytical look at the relationship between revenue generation model in the cyber-domain and how content producers of fake news benefit from it. This paper postulates that the existing online advertisement model remains lucrative for fake news publishers particularly in regions such as South Asia. Its interplay with social media platforms enables further monetisation and spreading of fake news to Internet users. It analyses the relationship between these factors and suggests cross-sector solutions for de-incentivising and countering fake news in the cyber domain. 

Can an online coaching programme facilitate behavioural change in women working in STEM fields?

With the communication and many aspects of social life moving into the online domain, social interaction and communication between individuals in corporate environments are often being transferred to the virtual realms in search of solutions to save time and money. This paper explores the way in which online coaching can provide a fruitful platform to introduce online communication and staff development in constructive ways where social and relational aspects of the group are accommodated.The two key areas of research presented in this paper are: (i) online coaching: exploring different modes of online coaching and the impact of combining individual and group coaching, and (ii) behavioural change: exploring different methods and approaches to develop self-confidence in young women.The sample chosen for this study is a group of young women. The reason this sample was chosen is that the author’s coaching practice has been delivering a three-month in-person coaching programme, focused on developing young women from different organisations, aged between mid-20s to mid 30s, for the last six years. With organisations cutting costs and individuals being short of time, the author wanted to investigate if an online coaching programme could offer an alternative solution to in-person coaching. Our research, backed up by the author’s own experience, also responds to the call for evidence and confirms previous claims (Followell, 2014) that if coaching is made available to women in the early stages of a career, this can help develop their self-confidence, self-belief and leadership skills.The research presented here is an exploratory, longitudinal study, using a sample size of 12 women, from two organisations, as well as one individual participant. The online coaching programme took place over six weeks, with quantitative and qualitative data gathered on two separate occasions, using questionnaires and semi-structured interviews. The data collected from the first questionnaire and semi-structured interview informed the design of the online coaching programme. The data collected from the second questionnaire and semi-structured interview measured the participant’s expectation against experience of online coaching. Thematic analysis was used to analyse the data.Drawing together current thinking, research objectives and findings, demonstrates how online coaching, using synchronous and asynchronous coaching, combined with individual and group coaching, creates a powerful range of methods to facilitate behavioural change. The study demonstrates that coaching young women, individually, as well as a group, helps to develop self-confidence. The programme also encouraged risk taking which led to new learning and behavioural change.

Solar Power Forecasting Based on Domain Adaptive Learning

Solar power forecasting is critical to ensure the safety and stability of the power grid with high photovoltaic power penetration. Machine learning methods are compelling in solar forecasting. These methods can capture the complex coupling relationship between different meteorological factors without physical modeling. Most of the existing machine learning based forecasts follow the batch learning manner. Once the training is completed, the structure and parameters of the model are usually no longer adjusted. However, the climate is complex and dynamic. It is difficult for a fixed model to adapt to the climate characteristics of different regions or periods. Therefore, an online domain adaptive learning approach is proposed in this paper. Knowledge can be selectively accumulated or forgotten in its iterative process. As weather changes, the model can dynamically adjust its structure to adapt to the latest weather conditions. Unlike existing adaptive iterative methods, the proposed adaptive learning approach does not rely on the labels of the test data in the updating process. Experiments show that this method can effectively track changes in data distribution and obtain reliable prediction results.