Bifrost

Cloud computing and big data: Technologies and applications

LinkedIn is the largest professional social network in the world with more than 238M members. It provides a platform for advertisers to reach out to professionals and target them using rich profile and behavioral data. Thus, online advertising is an important business for LinkedIn. In this talk, I will give an overview of machine learning and optimization components that power LinkedIn self-serve display advertising systems. The talk will not only focus on machine learning and optimization methods, but various practical challenges that arise when running such components in a real production environment. I will describe how we overcome some of these challenges to bridge the gap between theory and practice.The major components that will be described in details include Response prediction: The goal of this component is to estimate click-through rates (CTR) when an ad is shown to a user in a given context. Given the data sparseness due to low CTR for advertising applications in general and the curse of dimensionality, estimating such interactions is known to be a challenging. Furthermore, the goal of the system is to maximize expected revenue, hence this is an explore/exploit problem and not a supervised learning problem. Our approach takes recourse to supervised learning to reduce dimensionality and couples it with classical explore/exploit schemes to balance the explore/exploit tradeoff. In particular, we use a large scale logistic regression to estimate user and ad interactions. Such interactions are comprised of two additive terms a) stable interactions captured by using features for both users and ads whose coefficients change slowly over time, and b) ephemeral interactions that capture ad-specific residual idiosyncrasies that are missed by the stable component. Exploration is introduced via Thompson sampling on the ephemeral interactions (sample coefficients from the posterior distribution), since the stable part is estimated using large amounts of data and subject to very little statistical variance. Our model training pipeline estimates the stable part using a scatter and gather approach via the ADMM algorithm, ephemeral part is estimated more frequently by learning a per ad correction through an ad-specific logistic regression. Scoring thousands of ads at runtime under tight latency constraints is a formidable challenge when using such models, the talk will describe methods to scale such computations at runtime.Automatic Format Selection: The presentation of ads in a given slot on a page has a significant impact on how users interact with them. Web designers are adept at creating good formats to facilitate ad display but selecting the best among those automatically is a machine learning task. I will describe a machine learning approach we use to solve this problem. It is again an explore/exploit problem but the dimensionality of this problem is much less than the ad selection problem. I will also provide a detailed description of how we deal with issues like budget pacing, bid forecasting, supply forecasting and targeting. Throughout, the ML components will be illustrated with real examples from production and evaluation metrics would be reported from live tests. Offline metrics that can be useful in evaluating methods before launching them on live traffic will also be discussed.

Maintenance of embedded software systems is faced with multiple challenges, including the exploration and analysis of the actual system's runtime behavior. As a fundamental technique, tracing can be used to capture data about runtime behavior as a whole, and represents one of the few methods to observe and record data about embedded systems within their production environments. In this paper we present a software-based, function-boundary tracing approach for embedded software systems. It uses static binary instrumentation, which implies only lightweight memory and performance overheads. To further reduce these overheads, instrumentation can be configured per trace, i.e., activated only for a specified group of functions without having to recompile the system. The technique can be characterized by its robust implementation and its versatile usage. It is complemented by a visualization framework that allows for analysis and exploration of a system's runtime behavior, e.g., to examine thread interaction. To show the technique's applicability, we conclude with a case study that has been applied to an industrial embedded software system.

Continuous Delivery refers to a software development practice where members of a team frequently integrate their work, so that the process of delivery can be easily conducted. However, this continuous integration and delivery requires a reliable collaboration between development and IT operation teams. The DevOps practices support this collaboration since they enable that the operation staff making use of the same infrastructure as developers for their systems work. Our study aims at presenting a practical DevOps implementation and analyzing how the process of software delivery and infrastructure changes was automated. Our approach follows the principles of infrastructure as code, where a configuration platform – PowerShell DSC – was used to automatically define reliable environments for continuous software delivery. In this context, we defined the concept of “stage for dev”, also using the Docker technology, which involves all the elements that enable members of a team to have the same production environment, locally configured in their personal machines and thus empowering the continuous integration and delivery of system releases.

When a web service or application evolves, its database schema --- tables, constraints, and indices --- often need to evolve along with it. Depending on the database, some of these changes require a full table lock, preventing the service from accessing the tables under change. To deal with this, web services are typically taken offline momentarily to modify the database schema. However with the introduction of concepts like Continuous Deployment, web services are deployed into their production environments every time the source code is modified. Having to take the service offline --- potentially several times a day --- to perform schema changes is undesirable. In this paper we introduce QuantumDB--- a tool-supported approach that abstracts this evolution process away from the web service without locking tables. This allows us to redeploy a web service without needing to take it offline even when a database schema change is necessary. In addition QuantumDB puts no restrictions on the method of deployment, supports schema changes to multiple tables using changesets, and does not subvert foreign key constraints during the evolution process. We evaluate QuantumDB by applying 19 synthetic and 95 industrial evolution scenarios to our open source implementation of QuantumDB. These experiments demonstrate that QuantumDB realizes zerodowntime migrations at the cost of acceptable overhead, and is applicable in industrial continuous deployment contexts.

Cotton varieties develop differently because they present different physiological characteristics in production environments. This study aimed to know the physiological characteristics of cotton. The experiment was carried out at the São Paulo Agribusiness Technology Agency (APTA), Alta Paulista region. The experimental design used was Entirely Randomized (DIC), with six cotton cultivars: IMA5801B2RF; FM975WS; TMG47B2RF; TMG81WS; FM944GL and IACRDN. Where the following physiological parameters were determined: Assimilation rate CO2, leaf transpiration, stomatal conductance, internal concentration of CO2 in the substomatic chamber and efficient use of water. The cotton varieties showed different physiological characteristics, the IMA5801B2RF variety had the lowest performance, which may reflect low productivity. Further physiological studies are needed to understand the Cotton varieties behavior

As feature sizes get smaller the error of measurement must also become smaller. An adequate sub-micron measurement must be correct, within a certain or determined error limit. The error limit for a metrology instrument designed for use on sub-micron features in a production environment, the Opal 702, was investigated experimentally. There are numerous contributors to Iinewidth measurement uncertainty; (1) Dynamic Precision (including wafer and system variations), (2) system Stability (determined in a way to avoid including wafer stability), and (3) Offset (dependent on signal interpretation or algorithm) [1]. In a practical way all these variables are incorporated in (4) Dynamic Uncertainty.© (1990) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

The article deals with the formation of information and communication competence of junior high school students in computer science lessons in elementary school by means of project activities. The author emphasizes that the vocation of educational institutions in our time (including elementary school) consists in the formation of competences in the students of education, which today act as a guarantee of a successful professional and personal life of a person. Also, in the conditions of the rapid development of computer equipment and information technologies, the priority of education development is determined by the introduction of modern information and communication technologies in education, which ensure the improvement of the educational process, the accessibility and effectiveness of education, and the preparation of the young generation for life in the information society. And the most effective means of solving the tasks in our time are projects as an activity for solving various problems. Projects have certain features and significantly increase motivation to study, as they form a reasonable balance between academic knowledge and pragmatic skills. The use of project activities in the educational process makes it possible to comprehensively solve the tasks of education, upbringing and development of the school. The implementation of projects by younger schoolchildren makes it possible to form a personality that better adapts to changing conditions and circumstances, orients itself in various situations, successfully integrates into society, defines and implements its life strategy, is capable of self-organization, self-realization, and creative potential. An important role in solving project problems is played by information technologies, familiarization and assimilation of which takes place in computer science classes in elementary school. The use of information and communication technologies in project activities makes it possible to master new ways of working with alternative sources of knowledge, establish new contacts to expand the sphere of communication. The positive emotions that students receive in project activities using information and communication technologies strengthen their self-confidence, contribute to the development of their cognitive activity, and increase motivation to study school subjects. In this case, the implementation of projects using information and communication technologies should be an integral structural element of the educational process of primary school.

DevOps is a culture that emphasizes team cooperation and encourages enterprises to deliver software product features through automated processes. The Internet of Things paradigm is impulsively and continuously reforming industrial endeavours. DevOps combines continuous integration followed by continuous testing which is then succeeded by continuous deployment and continuous delivery. Continuous monitoring is a nerve ending once the product is delivered. This entire combination helps to produce faster deployments with quality. The DevOps-related challenges in IoT projects are discussed in this paper. The DevOps methodology and its recommended practices are frequently used in combination with version control, monitoring, and alerting at the device level. It has outlined the software architecture for IoT devices. This study reviewed research articles to investigate challenges with the Internet of Things applications related to DevOps framework. This study also tried to explore solutions for those challenges in form of guidelines, practices to adhere to, and models suggested in various available literature. The proposed article identifies important issues that must be resolved to implement the DevOps culture in IoT applications. Additionally, the organizations' cultural difficulties confronted while placing DevOps into practice are studied and a model is developed to provide guidelines. Effectiveness measures under consideration are collaboration, sharing, efficiency, and quality (Business value). So, this study emphasizes the challenges related to these effectiveness measures. According to research, there are several obstacles to implementing a DevOps culture in IoT applications, including poor communication, less no of DevOps experts, weak collaboration, and less management support. This research suggested a new model, Industrial DevOps Maturity Model (IDMM) to provide usage guidelines that can be the solution to existing challenges related to the effectiveness measures faced while using DevOps framework in IoT applications.

Experiments have proven that traditional prediction research methods have limitations in practice. Proposing countermeasures for environmental changes is the key to optimal control of the cold chain environment and reducing the lag of control effects. In this paper, a cold chain transportation environment prediction method, combining k-means++ and a long short-term memory (LSTM) neural network, is proposed according to the characteristics of the cold chain transportation environment of agricultural products. The proposed prediction model can predict the trend of cold chain environment changes in the next ten minutes, which allows cold chain vehicle managers to issue control instructions to the environmental control equipment in advance. The fusion process for temperature and humidity data measured by multiple data sensors is performed with the k-means++ algorithm, and then the fused data are fed into an LSTM neural network for prediction based on time series. The prediction error of the prediction model proposed in this paper is very satisfactory, with a root-mean-square error (RMSE), mean absolute error (MAE), mean squared error (MSE), mean absolute percentage error (MAPE) and R-squared of 0.5707, 0.2484, 0.3258, 0.0312 and 0.9660, respectively, for temperature prediction, and with an RMSE, MAE, MSE, mean absolute percentage error and R-squared of 1.6015, 1.1770, 2.5648, 0.2736 and 0.9702, respectively, for humidity prediction. Finally, the LSTM neural network and back propagation (BP) neural network are compared in order to enhance the reliability of the results. In terms of the prediction effect of the temperature and humidity in cold chain vehicles transporting agricultural products, the proposed model has a higher prediction accuracy than that of existing models and can provide strategic support for the fine management and regulation of the cold chain transportation environment.

Modern software development approaches rely extensively on tools. Motivated by practices such as continuous integration, deployment and delivery, these tools are used in a fashion where data are automatically accumulated in different databases as a side-effect of everyday development activities. In this paper we introduce an approach for software engineering data visualization as a mash up that combines data from issue management, software development and production use. The visualization can show to all stake holders how well continuous delivery is realized in the project. The visualization clearly shows the time spent to specify and develop the features as well the length of the delivery cycle. Further more the visualization shows how much work is unfinished and waiting for delivery. This can help the development team to decrease the amount of unfinished work and by that help them to keep up in continuous delivery mind set. In addition to development data usage of the features is also visualized.

Environmental characteristics of an invaded environment can predict the invasion success of a species, depending on its habits and life strategies. Cichla kelberi is a visual and voracious predator introduced in the Upper Paraná River floodplain, an area that suffers with several upstream dams that caused many environmental changes over the years (e.g. increased water transparency). As this species is a pre-adapted species to environments that presents high water transparency, our main goal was to test the hypothesis that variables related to the underwater visibility would be important drivers of the successful invasion of C. kelberi in the Upper Paraná River floodplain. We predict that turbidity (proxy of water transparency) is one of the major limnological determinants of occurrence and abundance of C. kelberi. Individuals of the invasive species and seven limnological variables were sampled quarterly between February 2000 and November 2018. This long-term data is part of the sixth site of the Brazilian Long Term Ecological Program (PELD). We fitted two regression models to all occurrence and abundance data (response variables) against the limnological variables (explanatory variables). For occurrence data, we fitted a Generalized Linear Mixed Model and for abundance data, we fitted a Linear Mixed Effects Model. The occurrence data of C. kelberi showed that turbidity and dissolved oxygen were the variables that most influenced the presence of this species, negatively and positively, respectively. The abundance data showed that dissolved oxygen and concentration of chlorophyll-a were the ones that most influenced the abundance of this species, negatively and positively, respectively. Our findings showed that besides turbidity, other limnological variables were also determinants for the occurrence and abundance of C. kelberi . Therefore, our results provided important information about the main environmental drivers of the establishment process of C. kelberi in the Upper Paraná River floodplain.

While continuous real-time software delivery practices induced by agile software development approaches create new business opportunities for organizations, these practices also present new security challenges in the DevOps environment. DevSecOps attempts to incorporate advanced automated security practices for agility in the DevOps environment. Mainstream perspectives of DevSecOps tend to overlook the collaborative role played by social actors and their relations with technologies in securing software applications in organizations. The first perspective emphasises the use of technologies such as containers, microservices, cryptographic protocols and origin authentication to secure the continuous deployment pipeline. The other dominant perspective focuses almost exclusively on the social aspects such as organizational silos, culture, and team collaboration. Such one-sided perspectives neglect the socio-technical argument that secure software applications from continuous deployment emerges when developers, quality assurers, operators and security experts combine their collective expertise together with DevSecOps technologies. The article presents a socio-technical framework of DevSecOps based on a systematic literature review. The review focused primarily, but not exclusively, on the computing and information systems literature and identified 26 peer reviewed articles from 2016 to 2020 which met the quality criteria and contributed to the analysis. The authors used a critical appraisal checklist and member checking to assess the quality of the articles. The authors then used thematic analysis to develop a comprehensive framework for DevSecOps based on the insights from these articles and a socio-technical lens. The socio-technical framework can be used by practitioners to perform a more holistic analysis of their DevSecOps practices. It highlights the key social and technical themes that underpin the effectiveness of DevSecOps and how insights about these themes can be used by practitioners to improve the instrumental and humanistic goals of DevSecOps. An interdisciplinary approach is proposed to adequately address challenging socio-technical relationships in DevSecOps. Future research can empirically test the importance of the interplay between technology and human activities to improve the overall performance of DevSecOps and other domains in cyber warfare and security.

Regardless of the size of organizations, they are migrating to microservice architecture for multiple reasons. Many of them have found that by embracing microservice architecture they can deliver software faster and efficiently. Managing small, fine-grained, and autonomous services is easier and allows teams to work on them independently. Adaptation of Agile practices together with methodologies such as the twelve-factor app for building microservices, enable flexibility, efficiency, and increase the speed of delivering the software. However, implementing a repeatable, adaptable, and reliable process for releasing software can be very challenging. Continuous Integration, Delivery, and Deployment workflows are determined by organization cultures and preferences as well as technical requirements. This paper presents an approach on how to design and implement a CI/CD process following the best practices on the build, test, deploy, and release processes. It also describes the challenges of designing CI/CD systems, and it outlines potential answers to those challenges.

Mamuka Zakaradze E-mail: mamuka.zakaradze@ciu.edu.ge PhD Students of Caucasus Internationl University Tbilisi, Georgia https://orcid.org/0009-0000-9022-5673 Abstract: The article aims to explore the impact of a multicultural environment on the decision-making process, both in a positive and negative context, and shed light on the diverse dynamics and challenges that arise when working groups interact with different cultures. In view of globalization, in the wake of the development of international relations in the field of other things, it is becoming more and more important to emphasize the importance of multiculturalism and the processes taking place in it, both in the creation of a fruitful working environment and in the decision-making process. Its role in terms of sharing experience, acquiring new knowledge and the possibility of innovation, which is accompanied by the challenges of multiculturalism in the process of creating an effective working environment, gaining trust in the team, adaptation and integration, should be emphasized here. Based on the research and analysis of the features of the multicultural environment, the paper contains conclusions and recommendations for the interested parties on what are the ways to create a productive and effective work environment in organizations characterized by a multicultural work team. Keywords: Multicultural environment, decision-making, teamwork, adaptation in a team, global challenges. JEL classification: M10; M12; M16.