Auditing Risk Prediction of Long-Term Unemployment

As more and more governments adopt algorithms to support bureaucratic decision-making processes, it becomes urgent to address issues of responsible use and accountability. We examine a contested public service algorithm used in Danish job placement for assessing an individual's risk of long-term unemployment. The study takes inspiration from cooperative audits and was carried out in dialogue with the Danish unemployment services agency. Our audit investigated the practical implementation of algorithms. We find (1) a divergence between the formal documentation and the model tuning code, (2) that the algorithmic model relies on subjectivity, namely the variable which focus on the individual's self-assessment of how long it will take before they get a job, (3) that the algorithm uses the variable "origin" to determine its predictions, and (4) that the documentation neglects to consider the implications of using variables indicating personal characteristics when predicting employment outcomes. We discuss the benefits and limitations of cooperative audits in a public sector context. We specifically focus on the importance of collaboration across different public actors when investigating the use of algorithms in the algorithmic society.

Future Movement Future – REJECTED

In a not-too-distant future, an anonymous researcher and their team applied for funding to develop their newest invention: a new algorithmic model for smart cameras that would allow people to analyze the movement of cars at a previously unheard-of scale. This system was said to enable new forms of predictive capabilities to emerge: the algorithm would be able to, for example, predict the route drivers wanted to take but had not yet taken—including, for example, their occult inner desires for getting away with a secret lover. A panel of academic reviewers from three different universities audited and reviewed the proposed system. All that is left are segments of the video-report resulting from this meeting, which became an urban legend among technology researchers. The short film “Future Movement Future – REJECTED” is the story of a dystopian surveillance future that was barred by institutional refusal. It importantly reminds us about how total surveillance, the “almighty algorithmic eye,” may end up seeing-predicting much less than imagining-dreaming.

Ergodic Algorithmic Model (EAM), with Water as Implicit Solvent, in Chemical, Biochemical, and Biological Processes

For many years, we have devoted our research to the study of the thermodynamic properties of hydrophobic hydration processes in water, and we have proposed the Ergodic Algorithmic Model (EAM) for maintaining the thermodynamic properties of any hydrophobic hydration reaction at a constant pressure from the experimental determination of an equilibrium constant (or other potential functions) as a function of temperature. The model has been successfully validated by the statistical analysis of the information elements provided by the EAM model for about fifty compounds. The binding functions are convoluted functions, RlnKeq = {f(1/T)* g(T)} and RTlnKeq = {f(T)* g(lnT)}, where the primary linear functions f(1/T) and f(T) are modified and transformed into parabolic curves by the secondary functions g(T) and g(lnT), respectively. Convoluted functions are consistent with biphasic dual-structure partition function, {DS-PF} = {M-PF} ∙ {T-PF} ∙ {ζw}, composed by ({M-PF} (Density Entropy), {T-PF}) (Intensity Entropy), and {ζw} (implicit solvent). In the present paper, after recalling the essential aspects of the model, we outline the importance of considering the solvent as “implicit” in chemical and biochemical reactions. Moreover, we compare the information obtained by computer simulations using the models till now proposed with “explicit” solvent, showing the mess of information lost without considering the experimental approach of the EAM model.

Forecasting Buying Intention Through Artificial Neural Network: An Algorithmic Solution on Direct-to-Consumer Brands

The direct-to-consumer (DTC) brands are emerging to reach more number of consumers with more potential to meet their expectations. They are characterized through their metamorphosis as the vertical brands sell their products from the manufacturer to consumers directly without any interruptions from distribution channels as in traditional mode of doing business. They are annihilating themselves in the virtual platforms and later disrupting their existing linear sales models. This empirical investigation is targeted to construct an algorithmic model through a deep learning process which has been instrumental to predict the purchase decision. This investigation has churned a predictive model that is based on the attributes of the buying behaviour of the consumers. The attributes of online buying behaviour like safety of transaction, availability of innovative products and quality of products have been considered to build a predictive model through artificial neural network (ANN). The accuracy of training and testing data are closer, which infers about the consistency and validity of the predictive model. There are several consequences arising from the predictive model obtained that can be seeded from customer-centred marketing and further stemmed from the framing of business strategy, gaining insights into market architecture and choice of customer

Analyzing Natural Digital Information in the Context of Market Research

The dynamics of irreversible multidimensional digitalization of production and consumption processes can be described today with a linear-positive or even exponential function. A significant part of the information background of a product, enterprise or brand is formed by their consumers, competitors or partners on the Internet, which considerably increases its accessibility and spread. Such kind of information can be called natural digital information (NDI). Its high market value is counterbalanced by its inhomogeneity and complexity for analysis. The solution to this problem lies in the field of creating automated tools for its subsequent search, aggregation, primary processing, quantification and analysis. The aim of this study is to describe the unique methodological properties of market research based on natural digital information. In order to achieve this aim, this study analyzes the theoretical basis in the field of NDI research, defines the categories of NDI and sources of its formation, describes the key properties of NDI, determines its advantages in comparison with other types of market information, and suggests a basic methodology for conducting typical NDI-based market research. An applied research study was carried out according to the designed methodology to show its advantages, as well as to describe the unique methodological properties of market research based on processing of NDI. The main result of this study is a universal algorithmic model for analyzing NDI in the context of market research, which includes a mechanism for defining and categorizing the digital sources of NDI, a model for forming the key properties of NDI, and basic classes of NDI analytical metrics. The toolkit developed by the authors allows market research to be conducted without direct attraction of research subjects, which results in cost reduction and elimination of the phenomenon of social desirability; this creates the so-called reasoned advertising messages that meet the requests of the target audience, which is proved by the big data that underlie the presented methodology. The developed algorithmic model is universal for analyzing natural digital information, and, with minor adaptations, can be used by any subject conducting market research.

Research on Intelligent Recognition Algorithm of Pneumonia Based on Deep Convolution and Attention Neural Network

Pneumonia is a common infection that inflames the air sacs in the lungs, causing symptoms such as difficulty breathing and fever. Although pneumonia is not difficult to treat, prompt diagnosis is crucial. Without proper treatment, pneumonia can be fatal, especially in children and the elderly. Chest x-rays are an affordable way to diagnose pneumonia. Investigating an algorithmic model that can reliably and intelligently classify pneumonia based on chest X-ray images could greatly reduce the burden on physicians. The advantages and disadvantages of each of the four convolutional neural networks VGG16, ResNet50, DenseNet201, and DWA algorithm models are analyzed and given by comparing and investigating each model. The VGG16, ResNet50, and DenseNet201 network models are compared with the DWA model. When training the depthwise separable convolution with attention neural network (DWA), the training accuracy reaches 97.5%. The validation accuracy was 79% due to the model’s tendency to overfit, and the test dataset had 1175 X-ray images with a test accuracy of 96.1%. The experimental results illustrate the effectiveness of the attention mechanism and the reliability of the deeply separable convolutional neural network algorithm. The successful application of the deep learning algorithm proposed in this paper on pneumonia recognition will provide an objective, accurate, and fast solution for medical practitioners and can provide a fast and accurate pneumonia diagnosis system for doctors.

Algorithmic model of social processes

The development of the social sciences needs to rely on precise methods. The nomological model of explanation adopted in the natural sciences is ill-suited for the social sciences. An algorithmic model of society can be a promising solution to existing problems. In its most general form, an algorithm is a generally understood prescription for what actions to perform and in what order to achieve the desired result. Any algorithm can be represented as a set of rules of the form «If A, do D to get P». People are the bearers of this kind of rules that apply in different areas of their activities. The rules are subject to change based on personal and collective experience. There is a special mathematical discipline that studies the laws of evolution of such rules. This discipline is called genetic (evolutionary) programming. Contrary to the threatening name, the algorithmic model does not imply the deprivation of a person’s right to free choice, but it needs this right as a necessary condition for the evolution of social algorithms. These algorithms allow us to give a non-causal, but law-like explanation of many well-known social phenomena, as well as to effectively model the future, which is critically important today. A retrospective look at the evolution of social algorithms shows that the current global crisis of human society is associated with the approach to the point of singularity in their evolution. This is due to the fact that there is no need for direct human participation in the implementation of social algorithms, which is reflected in a fundamental change in the sphere of employment and less need for further development of the sciences.

A web-based model to enhance algorithmic thinking for novice programmers

Algorithmic thinking is considered as one of the important steps toward learning to code for novices in programming education. In this study, algorithmic thinking was promoted by introducing a Problem Analysis Algorithmic Model (PAAM) in an Algorithms and Programming 1 (AP) course. A web-based application is developed to offer the PAAM model in the course. The application includes all teaching topics taught in the AP course. One-way cluster sampling and quantitative research were used in this research study. The impact of the PAAM model on novices was determined by conducting a survey. t-test was performed to analyze the students’ responses. The final exam grades for the last two semesters were compared to examine the effect of the PAAM model on students’ gain in the AP course. Results show that novice programmers appreciated the PAAM model in the AP course teaching processes and learning activities. The model supports novices to understand the programming question requirements (input, process, and output) and promotes algorithmic thinking. Moreover, the model helps students in learning problem-solving skills, understanding programming concepts and structures. It also focuses on students’ cognitive engagement and gain in programming. The model not only impacts positively on students’ gain but it also helps in reducing the attrition rates (failure and dropout) in the course.