Refine
Document Type
- Master's Thesis (55) (remove)
Has Fulltext
- yes (55)
Is part of the Bibliography
- no (55) (remove)
Keywords
- IT-Sicherheit (6)
- Maschinelles Lernen (4)
- Computersicherheit (3)
- Deep learning (3)
- E-Learning (3)
- Künstliche Intelligenz (3)
- Automatisierung (2)
- Change Management (2)
- Digitalisierung (2)
- E-Commerce (2)
Institute
- Fakultät Medien (M) (ab 22.04.2021) (55) (remove)
Open Access
- Closed (35)
- Closed Access (17)
- Open Access (3)
- Diamond (2)
Linux and Linux-based operating systems have been gaining more popularity among the general users and among developers. Many big enterprises and large companies are using Linux for servers that host their websites, some even require their developers to have knowledge about Linux OS. Even in embedded systems one can find many Linux-based OS that run them. With its increasing popularity, one can deduce the need to secure such a system that many personnel rely on, be it to protect the data that it stores or to protect the integrity of the system itself, or even to protect the availability of the services it offers. Many researchers and Linux enthusiasts have been coming up with various ways to secure Linux OS, however new vulnerabilities and new bugs are always found, by malicious attackers, with every update or change, which calls for the need of more ways to secure these systems.
This Thesis explores the possibility and feasibility of another way to secure Linux OS, specifically securing the terminal of such OS, by altering the commands of the terminal, getting in the way of attackers that have gained terminal access and delaying, giving more time for the response teams and for forensics to stop the attack, minimize the damage, restore operations, and to identify collect and store evidence of the cyber-attack. This research will discuss the advantages and disadvantages of various security measures and compare and contrast with the method suggested in this research.
This research is significant because it paints a better picture of what the state of the art of Linux and Linux-based operating systems security looks like, and it addresses the concerns of security enthusiasts, while exploring new uncharted area of security that have been looked at as a not so significant part of protecting the OSes out of concern of the various limitations and problems it entails. This research will address these concerns while exploring few ways to solve them, as well as addressing the ideal areas and situations in which the proposed method can be used, and when would such method be more of a burden than help if used.
On a regular basis, we hear of well-known online services that have been abused or compromised as a result of data theft. Because insecure applications jeopardize users' privacy as well as the reputation of corporations and organizations, they must be effectively secured from the outset of the development process. The limited expertise and experience of involved parties, such as web developers, is frequently cited as a cause of risky programs. Consequently, they rarely have a full picture of the security-related decisions that must be made, nor do they understand how these decisions affect implementation accurately.
The selection of tools and procedures that can best assist a certain situation in order to protect an application against vulnerabilities is a critical decision. Regardless of the level of security that results from adhering to security standards, these factors inadvertently result in web applications that are insufficiently secured. JavaScript is a language that is heavily relied on as a mainstream programming language for web applications with several new JavaScript frameworks being released every year.
JavaScript is used on both the server-side in web applications development and the client-side in web browsers as well.
However, JavaScript web programming is based on a programming style in which the application developer can, and frequently must, automatically integrate various bits of code from third parties. This potent combination has resulted in a situation today where security issues are frequently exploited. These vulnerabilities can compromise an entire server if left unchecked. Even though there are numerous ad hoc security solutions for web browsers, client-side attacks are also popular. The issue is significantly worse on the server side because the security technologies available for server-side JavaScript application frameworks are nearly non-existent.
Consequently, this thesis focuses on the server-side aspect of JavaScript; the development and evaluation of robust server-side security technologies for JavaScript web applications. There is a clear need for robust security technologies and security best practices in server-side JavaScript that allow fine-grained security.
However, more than ever, there is this requirement of reducing the associated risks without hindering the web application in its functionality.
This is the problem that will be tackled in this thesis: the development of secure security practices and robust security technologies for JavaScript web applications, specifically, on the server-side, that offer adequate security guarantees without putting too many constraints on their functionality.
Though the basic concept of a ledger that anyone can view and verify has been around for quite some time, today’s blockchains bring much more to the table including a way to incentivize users. The coins given to the miner or validator were the first source of such incentive to make sure they fulfilled their duties. This thesis draws inspiration from other peer efforts and uses this same incentive to achieve certain goals. Primarily one where users are incentivised to discuss their opinions and find scientific or logical backing for their standpoint. While traditional chains form a consensus on a version of financial "truth", the same can be applied to ideological truths too. To achieve this, creating a modified or scaled proof of stake consensus mechanism is explored in this work. This new consensus mechanism is a Reputation Scaled - Proof of Stake. This reputation can be built over time by voting for the winning side consistently or by sticking to one’s beliefs strongly. The thesis hopes to bridge the gap in current consensus algorithms and incentivize critical reasoning.
The progress in machine learning has led to advanced deep neural networks. These networks are widely used in computer vision tasks and safety-critical applications. The automotive industry, in particular, has experienced a significant transformation with the integration of deep learning techniques and neural networks. This integration contributes to the realization of autonomous driving systems. Object detection is a crucial element in autonomous driving. It contributes to vehicular safety and operational efficiency. This technology allows vehicles to perceive and identify their surroundings. It detects objects like pedestrians, vehicles, road signs, and obstacles. Object detection has evolved from being a conceptual necessity to an integral part of advanced driver assistance systems (ADAS) and the foundation of autonomous driving technologies. These advancements enable vehicles to make real-time decisions based on their understanding of the environment, improving safety and driving experiences. However, the increasing reliance on deep neural networks for object detection and autonomous driving has brought attention to potential vulnerabilities within these systems. Recent research has highlighted the susceptibility of these systems to adversarial attacks. Adversarial attacks are well-designed inputs that exploit weaknesses in the deep learning models underlying object detection. Successful attacks can cause misclassifications and critical errors, posing a significant threat to the functionality and safety of autonomous vehicles. With the rapid development of object detection systems, the vulnerability to adversarial attacks has become a major concern. These attacks manipulate inputs to deceive the target system, significantly compromising the reliability and safety of autonomous vehicles. In this study, we focus on analyzing adversarial attacks on state-of-the-art object detection models. We create adversarial examples to test the models’ robustness. We also check if the attacks work on a different object detection model meant for similar tasks. Additionally, we extensively evaluate recent defense mechanisms to see how effective they are in protecting deep neural networks (DNNs) from adversarial attacks and provide a comprehensive overview of the most commonly used defense strategies against adversarial attacks, highlighting how they can be implemented practically in real-world situations.
Much of the research in the field of audio-based machine learning has focused on recreating human speech via feature extraction and imitation, known as deepfakes. The current state of affairs has prompted a look into other areas, such as the recognition of recording devices, and potentially speakers, by only analysing sound files. Segregation and feature extraction are at the core of this approach.
This research focuses on determining whether a recorded sound can reveal the recording device with which it was captured. Each specific microphone manufacturer and model, among other characteristics and imperfections, can have subtle but compounding effects on the results, whether it be differences in noise, or the recording tempo and sensitivity of the microphone while recording. By studying these slight perturbations, it was found to be possible to distinguish between microphones based on the sounds they recorded.
After the recording, pre-processing, and feature extraction phases we completed, the prepared data was fed into several different machine learning algorithms, with results ranging from 70% to 100% accuracy, showing Multi-Layer Perceptron and Logistic Regression to be the most effective for this type of task.
This was further extended to be able to tell the difference between two microphones of the same make and model. Achieving the identification of identical models of a microphone suggests that the small deviations in their manufacturing process are enough of a factor to uniquely distinguish them and potentially target individuals using them. This however does not take into account any form of compression applied to the sound files, as that may alter or degrade some or most of the distinguishing features that are necessary for this experiment.
Building on top of prior research in the area, such as by Das et al. in in which different acoustic features were explored and assessed on their ability to be used to uniquely fingerprint smartphones, more concrete results along with the methodology by which they were achieved are published in this project’s publicly accessible code repository.
Entwicklung eines Kommunikationskonzepts für Black Forest Formula an der Hochschule Offenburg
(2021)
Die vorliegende Masterthesis befasst sich mit der Entwicklung eines Kommunikationskonzepts für das Black Forest Formula Projekt an der Hochschule Offenburg. Die bisherigen kommunikativen Aktivitäten werden im Rahmen dieser Arbeit überarbeitet und erweitert, sodass ein konsistentes Gesamtkonzept für die Kommunikation entsteht. Hierzu werden die einzelnen Phasen eines Konzeptionsprozesses durchlaufen. Das Ziel ist eine langfristige Planung, welche die Interessen der verschiedenen Gruppen innerhalb und außerhalb des Projekts berücksichtigt, um deren Unterstützung zu sichern. Des Weiteren wird durch die Ausrichtung auf festgelegte Kommunikationsziele eine Verbesserung der internen und externen Kommunikation von Black Forest Formula angestrebt. Die Arbeit richtet sich daher insbesondere an die leitenden Mitglieder des Teams und soll diesen als Handlungsempfehlung dienen. Darüber hinaus ist die Arbeit auch für all jene interessant, die sich mit dem Ablauf und den Methoden zur Entwicklung von Kommunikationskonzepten befassen möchten.
The identification of vulnerabilities is an important element of the software development process to ensure the security of software. Vulnerability identification based on the source code is a well studied field. To find vulnerabilities on the basis of a binary executable without the corresponding source code is more challenging. Recent research has shown how such detection can be performed statically and thus runtime efficiently by using deep learning methods for certain types of vulnerabilities.
This thesis aims to examine to what extent this identification can be applied sufficiently for a variety of vulnerabilities. Therefore, a supervised deep learning approach using recurrent neural networks for the application of vulnerability detection based on binary executables is used. For this purpose, a dataset with 50,651 samples of 23 different vulnerabilities in the form of a standardised LLVM Intermediate Representation was prepared. The vectorised features of a Word2Vec model were then used to train different variations of three basic architectures of recurrent neural networks (GRU, LSTM, SRNN). For this purpose, a binary classification was trained for the presence of an arbitrary vulnerability, and a multi-class model was trained for the identification of the exact vulnerability, which achieved an out-of-sample accuracy of 88% and 77%, respectively. Differences in the detection of different vulnerabilities were also observed, with non-vulnerable samples being detected with a particularly high precision of over 98%. Thus, the methodology presented allows an accurate detection of vulnerabilities, as well as a strong limitation of the analysis scope for further analysis steps.
Cloud computing is a combination of technologies, including grid computing and distributed computing, that use the Internet as a network for service delivery. Organizations can select the price and service models that best accommodate their demands and financial restrictions. Cloud service providers choose the pricing model for their cloud services, taking the size, usage, user, infrastructure, and service size into account. Thus, cloud computing’s economic and business advantages are driving firms to shift more applications to the cloud, boosting future development. It enlarges the possibilities of current IT systems.
Over the past several years, the ”cloud computing” industry has exploded in popularity, going from a promising business concept to one of the fastest expanding areas of the IT sector. Most enterprises are hosting or installing web services in a cloud architecture for management simplicity and improved availability. Virtual environments are applied to accomplish multi-tenancy in the cloud. A vulnerability in a cloud computing environment poses a direct threat to the users’ privacy and security. In our digital age, the user has many identities. At all levels, access rights and digital identities must be regulated and controlled.
Identity and access management(IAM) are the process of managing identities and regulating access privileges. It is considered as a front-line soldier of IT security. It is the goal of identity and access management systems to protect an organization’s assets by limiting access to just those who need it and in the appropriate cases. It is required for all businesses with thousands of users and is the best practice for ensuring user access control. It identifies, authenticates, and authorizes people to access an organization’s resources. This, in turn, enhances access management efficiency. Authentication, authorization, data protection, and accountability are just a few of the areas in which cloud-based web services have security issues. These features come under identity and access management.
The implementation of identity and access management(IAM) is essential for any business. It’s becoming more and more business-centric, so we need more than technical know-how to succeed. Organizations may save money on identity management and, more crucially, become much nimbler in their support of new business initiatives if they have developed sophisticated IAM capabilities. We used these features of identity and access management to validate the robustness of the cloud computing environment with a comparison of traditional identity and access management.
Die rasante Digitalisierung verändert die Hochschule Offenburg nachhaltig. Jedes Semester entstehen zahlreiche akademische Arbeiten und Prüfungsdaten, die von hoher Bedeutung für die Qualitätssicherung und den Bildungsprozess sind. Bisher fehlte jedoch eine effiziente Lösung zur Archivierung dieser Daten. Meine Masterarbeit präsentiert ein neues Archivierungskonzept, das eine zentrale, digitale Plattform schafft. Diese ermöglicht es Lehrenden und Studierenden, leicht auf ihre Daten zuzugreifen, sei es für Projektarbeiten, Seminarleistungen oder Prüfungen. Ich stelle nicht nur das Konzept vor, sondern auch die Umsetzung eines Prototyps auf Intrexx, einer Low-Code-Entwicklungsplattform. Mein Ziel ist es, ein Handbuch für zukünftige Entwickler zu hinterlassen, um zur digitalen Transformation der Hochschule beizutragen und die Bildungsprozesse zu optimieren.
In this work, an implementation of the somewhat homomorphic BV encryption scheme is presented. During the implementation, care was taken to ensure that the resulting program will be as efficient as possible i.e. fast and resource-saving. The basis for this is the work of Arndt Bieberstein, who implemented the BV scheme with respect to functionality. The presented implementation supports the basics of the BV scheme, namely (symmetric and asymmetric) encryption, decryption and evaluation of addition as well as multiplication. Additionally, it supports the encoding of positive and negative numbers, various gaussian sampling methods, basically infinitely large polynomial coefficients, the generation of suitable parameters for a use case, threading and relinearization to reduce the size of a ciphertext after multiplications. After presenting the techniques used in the implementation, it’s actual efficiency is determined by measuring the timings of the operations for various parameters.
Die vorliegende Arbeit beschäftigt sich mit der Nutzung von Reinforcement Learning in der Informationsbeschaffungs-Phase eines Penetration Tests. Es werden Kernprobleme in den bisherigen Ansätzen anderer das Thema betreffender wissenschaftlicher Arbeiten analysiert und praktische Lösungsansätze für diese bisherigen Hindernisse vorgestellt und implementiert. Die Arbeit zeigt damit eine beispielhafte Implementierung eines Reinforcement Learning Agenten zur Automatisierung der Informationsbeschaffungs-Phase eines Penetration Tests und stellt Lösungen für existierende Probleme in diesem Bereich dar.
Eingebettet wird diese wissenschaftliche Arbeit in die Anforderungen der Herrenknecht AG hinsichtlich der Absicherung des Tunnelbohrmaschinen-Netzwerks. Dabei werden praktische Ergebnisse des eigen entwickelten Reinforcement Learning Modells im Tunnelbohrmaschinen-Test-Netzwerk der Herrenknecht AG vorgestellt.
Empfehlungssysteme sind auf E-Commerce-Webseiten omnipräsent und unterstützen die Nutzer an bestimmten Touchpoints beim Onlineshopping, indem sie auf Produkte aufmerksam machen. In den meisten Anwendungsfällen werden Produkte empfohlen, die den Interessen der Nutzer entsprechen oder einen Warenkorb komplettieren sollen. Während klassische Empfehlungssysteme die Nutzer meistens zu Beginn oder zum Ende einer Produktannäherung erreichen, finden sie dazwischen selten Anwendung. Dabei könnten Empfehlungssysteme bereits in die Navigations-Journey der Nutzer integriert werden und so maßgeblich zur Produktauffindbarkeit beitragen. Trotz der Tatsache, dass Empfehlungssysteme maßgeblich zum Geschäftserfolg im Online-Handel beitragen, ist ihre Integration in die Navigation von E-Commerce-Plattformen noch wenig erforscht.
Ziel dieser Arbeit ist es daher zu ergründen, wie ein Empfehlungssystem in der lokalen Navigation von E-Commerce-Plattformen ausgestaltet sein sollte. Dabei soll der Prototyp eines Empfehlungssystems in der lokalen Navigation konzipiert werden, indem einerseits echte Warenkorbdaten einer E-Commerce-Plattform und andererseits Nutzerpräferenzen untersucht werden. Die Untersuchung umfasst die algorithmische Datenverarbeitung von Warenkorbdaten und die Durchführung einer quantitativen Befragung. Darüber hinaus wurde domänenspezifisches Wissen über die Verfahren, Algorithmen, Designkriterien und weitere Erfolgsfaktoren recherchiert.
Die Ergebnisse deuten darauf hin, dass die alleinige Verarbeitung von Warenkorbdaten nicht ausreichend für ein navigatorisches Empfehlungssystem ist, obwohl sich die Alternating Least Squares Matrixfaktorisierung als geeignetes Verfahren herausgestellt hat. Des Weiteren können Empfehlungssysteme in der lokalen Navigation sowohl einen positiven als auch einen negativen Einfluss auf die Nutzerzufriedenheit haben. Aus den Ergebnissen konnte ein detaillierter Prototyp konzipiert und vorgestellt werden. Bei diesem Prototyp handelt es sich um ein Session-basiertes Empfehlungssystem, das den Onlineshopping-Kontext der Nutzer ermitteln kann. Auf dieser Basis kann das System Produktkategorien in Echtzeit empfehlen, die zu diesem Kontext komplementär sind oder einen Social Proof abbilden.
It is generally agreed that the development and deployment of an important amount of IoT devices throughout the world has revolutionized our lives in a way that we can rely on these devices to complete certain tasks that may have not been possible just years ago which also brought a new level of convenience and value to our lives.
This technology is allowing us in a smart home environment to remotely control doors, windows, and fridges, purchase online, stream music easily with the use of voice assistants such as Amazon Echo Alexa, also close a garage door from anywhere in the world to cite some examples as this technology has added value to several domains ranging from household environments, cites, industries by exchanging and transferring data between these devices and customers. Many of these devices’ sensors, collect and share information in real-time which enables us to make important business decisions.
However, these devices pose some risks and also some security and privacy challenges that need to be addressed to reach their full potential or be considered to be secure. That is why, comprehensive risk analysis techniques are essential to enhance the security posture of IoT devices as they can help evaluate the robustness and reliability towards potential susceptibility to risks, and vulnerabilities that IoT devices in a smart home setting might possess.
This approach relies on the basis of ISO/IEC 27005 methodology and risk matrix method to highlight the level of risks, impact, and likelihood that an IoT device in smart home settings can have, map the related vulnerability, threats and risks and propose the necessary mitigation strategies or countermeasures that can be taken to secure a device and therefore satisfying some security principles. Around 30 risks were identified on Amazon Echo and the related IoT system using the methodology. A detailed list of countermeasures is proposed as a result of the risk analysis. These results, in turn, can be used to elevate the security posture of the device.
Among the billions of smartphone users in the world, Android still holds more than 80% of the market share. The applications which the users install have a specific set of features that need access to some device functionalities and sensors that may hold sensitive information about the user. Therefore, Android releases have set permission standards to let the user know what information is being disclosed to the application. Along with other security and privacy improvements, significant changes to the permission scheme are introduced with the Android 6.0 version (API level 23). In this master thesis, the Android permission scheme is tested on two devices from different eras. The evolution of Android over the years is examined in terms of confidentiality. For each device, two applications are built; one focused on extracting every piece of information within the confidentiality scope with every permission declared and/or requested, and the other app focused on getting this type of information without user notification. The resulting analysis illustrates whether how and in what way the Android permission scheme declined or improved over time.
Die vorliegende Masterarbeit setzt sich mit dem Thema „Digitale Medien und soziale Anerkennung“ auseinander. Dabei wird der Frage nachgegangen, welchen Einfluss die Nutzung digitaler Medien auf das Selbstbild junger Erwachsener, auf das Thema Cybermobbing und den zwischenmenschlichen Umgang in der Gesellschaft hat. Als Praxisbeispiel wurde exemplarisch die Folge „Nosedive“ der dystopischen Sci-Fi-Serie „Black Mirror“ ausgewählt, in der eine Gesellschaft porträtiert wird, deren Leben durch ein soziales Bewertungssystem bestimmt wird. Im Zuge dessen soll ein Bezug zu realen Sozialkredit-Systemen hergestellt werden. Das Ziel der Forschung ist es, eine umfangreiche Analyse der Auswirkungen des Medienkonsums von Fiktion und Realität vor dem Hintergrund sozialer Anerkennung durchzuführen.
Im theoretischen Teil der Arbeit soll allgemein auf digitale Medien, das Nutzungsverhalten von Jugendlichen sowie auf die Bedeutung von Identität und Selbstbild (im Netz) eingegangen werden. Sowohl die Einflüsse der digitalen Mediennutzung auf die Identität, das Selbstbild sowie die Selbstdarstellung als auch die allgemeinen positiven und negativen Auswirkungen durch exzessiven Medienkonsum von Jugendlichen sollen in dieser Arbeit näher beleuchtet und einander gegenübergestellt werden. Im nächsten Schritt sollen die Auswirkungen auf die Bereiche Cybermobbing sowie die zwischenmenschliche Kommunikation und soziale Beziehungen untersucht werden. Nach Abhandlung der Einflüsse und Auswirkungen auf die Jugendlichen unserer Gesellschaft folgt der praktische Teil der Arbeit, in dem zunächst die Anthologie-Serie „Black Mirror“ und die Inhalte der Folge „Nosedive“ vorgestellt werden sollen. Anhand ausgewählter Merkmale soll die fiktionale Welt mit unserer modernen Gesellschaft analysiert und verglichen werden. Abschließend soll auf die Sozialkredit-Systeme in China eingegangen und ein Bezug zur Folge hergestellt werden. In der Analyse sollen die Ergebnisse der vorangegangenen Kapitel miteinbezogen werden. Die Ergebnisse der Arbeit zeigen, dass digitale Medien einen weitreichenden sowohl positiven als auch negativen Einfluss auf Jugendliche haben. Die Nutzung hat viele Bereiche unseres Lebens bereits verändert. Besonders soziale Medien hinterlassen ihre Spuren in der Identität und Werteentwicklung von Kindern und Jugendlichen. Obwohl sie in der jüngeren Generation hohes Ansehen genießen, bestimmen sie jedoch (noch) nicht über andere Bereiche des Lebens.
Even though the internet has only been there for a short period, it has grown tremendously. To- day, a significant portion of commerce is conducted entirely online because of increased inter- net users and technological advancements in web construction. Additionally, cyberattacks and threats have expanded significantly, leading to financial losses, privacy breaches, identity theft, a decrease in customers’ confidence in online banking and e-commerce, and a decrease in brand reputation and trust. When an attacker pretends to be a genuine and trustworthy institution, they can steal private and confidential information from a victim. Aside from that, phishing has been an ongoing issue for a long time. Billions of dollars have been shed on the global economy. In recent years, there has been significant progress in the development of phishing detection and identification systems to protect against phishing attacks. Phishing detection technologies frequently produce binary results, i.e., whether a phishing attempt was made or not, with no explanation. On the other hand, phishing identification methodologies identify phishing web- pages by visually comparing webpages with predetermined authentic references and reporting phishing together with its target brand, resulting in findings that are understandable. However, technical difficulties in the field of visual analysis limit the applicability of currently available solutions, preventing them from being both effective (with high accuracy) and efficient (with little runtime overhead). Here, we evaluate existed framework called Phishpedia. This hybrid deep learning system can recognize identity logos from webpage screenshots and match logo variants of the same brand with high precision. Phishpedia provides high accuracy with low run- time. Lastly, unlike other methods, Phishpedia does not require training on any phishing sam- ples whatsoever. Phishpedia exceeds baseline identification techniques (EMD, PhishZoo, and LogoSENSE), inaccurately detecting phishing pages in lengthy testing using accurate phishing data. The effectiveness of Phishpedia was tested and compared against other standard machine learning algorithms and some state-of-the-art algorithms. The given solutions performed better than different algorithms in the given dataset, which is impressive.
Privacy is the capacity to keep some things private despite their social repercussions. It relates to a person’s capacity to control the amount, time, and circumstances under which they disclose sensitive personal information, such as a person’s physiology, psychology, or intelligence. In the age of data exploitation, privacy has become even more crucial. Our privacy is now more threatened than it was 20 years ago, outside of science and technology, due to the way data and technology highly used. Both the kinds and amounts of information about us and the methods for tracking and identifying us have grown a lot in recent years. It is a known security concern that human and machine systems face privacy threats. There are various disagreements over privacy and security; every person and group has a unique perspective on how the two are related. Even though 79% of the study’s results showed that legal or compliance issues were more important, 53% of the survey team thought that privacy and security were two separate things. Data security and privacy are interconnected, despite their distinctions. Data security and data privacy are linked with each other; both are necessary for the other to exist. Data may be physically kept anywhere, on our computers or in the cloud, but only humans have authority over it. Machine learning has been used to solve the problem for our easy solution. We are linked to our data. Protect against attackers by protecting data, which also protects privacy. Attackers commonly utilize both mechanical systems and social engineering techniques to enter a target network. The vulnerability of this form of attack rests not only in the technology but also in the human users, making it extremely difficult to fight against. The best option to secure privacy is to combine humans and machines in the form of a Human Firewall and a Machine Firewall. A cryptographic route like Tor is a superior choice for discouraging attackers from trying to access our system and protecting the privacy of our data There is a case study of privacy and security issues in this thesis. The problems and different kinds of attacks on people and machines will then be briefly talked about. We will explain how Human Firewalls and machine learning on the Tor network protect our privacy from attacks such as social engineering and attacks on mechanical systems. As a real-world test, we will use genomic data to try out a privacy attack called the Membership Inference Attack (MIA). We’ll show Machine Firewall as a way to protect ourselves, and then we’ll use Differential Privacy (DP), which has already been done. We applied the method of Lasso and convolutional neural networks (CNN), which are both popular machine learning models, as the target models. Our findings demonstrate a logarithmic link between the desired model accuracy and the privacy budget.
Licht war für die Menschheit schon immer ein Hilfsmittel zur Orientierung. Das Zusammenspiel zwischen hellen und schattierten Oberflächen macht eine räumliche Wahrnehmung erst möglich. Die Lokalisierung von Lichtquellen bietet darüber hinaus für zahlreiche Anwendungsfelder, wie beispielsweise Augmented Reality, ein großes Potential.
Das Ziel der vorliegenden Arbeit war es, ein neuronales Netzwerk zu entwickeln, welches mit Hilfe eines selbst generierten, synthetischen Datensatzes eine Lichtsetzung parametrisiert. Dafür wurden State-of-the-Art Netzwerke aus der digitalen Bildverarbeitung eingesetzt.
Zu Beginn der Arbeit mussten die Eigenschaften der Lichtsetzung extrahiert werden. Eine weitere fundamentale Anforderung war die Aufbereitung des Wissens von Deep Learning.
Für die Generierung des synthetischen Datensatzes wurde eigens ein Framework entwickelt, welches auf der Blender Engine basiert.
Anschließend wurden die generierten Bilder und Metadaten in einem abgewandelten VGG16- und ResNet50-Netz trainiert, validiert und evaluiert.
Eine gewonnene Erkenntnis ist, dass sich künstlich generierte Daten eignen um ein neuronales Netz zu trainieren. Des Weiteren konnte gezeigt werden, dass sich mit Hilfe von Deep Learning Lichtsetzungsparameter extrahieren lassen.
Eine weiterführende Forschungsaufgabe könnte mit dem vorgeschlagenen Ansatzdie Lichtinszenierung von Augmented Reality Anwendungen verbessern.
Für Unternehmen ist es zunehmend von Interesse, durch Social-Media-Kommunikation nicht nur die Aufmerksamkeit der Zielgruppe zu wecken, sondern mit der aufmerksamkeitsstarken Ansprache die Wahrnehmung ihrer Marke und das marktbezogene Verhalten zu beeinflussen. Humorvolle Appelle sind in diesem Zusammenhang zur gängigen Werbepraxis geworden und finden auf Social Media in der direkten Interaktion zwischen Marken und ihrer Zielgruppe statt. Mit der vorliegenden Arbeit wird humorvoll-schlagfertige Unternehmenskommunikation auf Social Media untersucht. Das Ziel war es, die Wirkungszusammenhänge zwischen derartiger Kommunikation, deren Bewertung, dem Markenimage, der Markenauthentizität und den Handlungsabsichten besser zu verstehen. Die Erkenntnisse über diese Zusammenhänge können sowohl der weiteren Erforschung als auch künftigen Marketingentscheidungen dienen.
Im Zuge dessen wurden die theoretischen Hintergründe moderner Markenkommunikation sowie der Konzepte Humor und Schlagfertigkeit ausgearbeitet. Der empirische Teil der Arbeit besteht aus der Konzipierung, Durchführung, Analyse und Ergebnisdiskussion einer quantitativen Studie. Zu den zentralen Ergebnissen zählt, dass sich der Grad an Humor und an Schlagfertigkeit darauf auswirkt, wie sehr die Kommunikation gefällt. Wenn diese gut gefällt, stärkt das wiederum das Markenimage. Auch die wahrgenommene Markenauthentizität bestimmt das Markenimage und wird selbst vor allem durch den Grad der Schlagfertigkeit bestimmt. Das Markenimage beeinflusst die Weiterleitungsabsicht der Kommunikation (virale Effekte) und die Kaufabsicht. Dabei unterscheidet sich humorvoll-schlagfertige Unternehmenskommunikation von neutraler Vergleichskommunikation hinsichtlich der Weiterleitungsabsicht signifikant.