arXiv subjects

Daniel Otten

Publications and source records attributed to Daniel Otten.

13 recordsLinked to original sources

arXiv2026

"Don't Be Afraid, Just Learn": Insights from Industry Practitioners to Prepare Software Engineers in the Age of Generative AI

Although tension between university curricula and industry expectations has existed in some form for decades, the rapid integration of generative AI (GenAI) tools into software development has recently widened the gap between the two domains. To better understand this disconnect, we surveyed 51 industry practitioners (software developers, technical leads, upper management, \etc) and conducted 11 follow-up interviews focused on hiring practices, required job skills, perceived shortcomings in university curricula, and views on how university learning outcomes can be improved. Our results suggest that GenAI creates demand for new skills (\eg prompting and output evaluation), while strengthening the importance of soft-skills (\eg problem solving and critical thinking) and traditional competencies (\eg architecture design and debugging). We synthesize these findings into actionable recommendations for academia (\eg how to incorporate GenAI into curricula and evaluation redesign). Our work offers empirical guidance to help educators prepare students for modern software engineering environments.

cs.SE↗

arXiv2026

No Traffic to Cry: Traffic-Oblivious Link Deactivation for Green Traffic Engineering

As internet traffic grows, the underlying infrastructure consumes increasing amounts of energy. During off-peak hours, large parts of the networks remain underutilized, presenting significant potential for energy savings. Existing Green Traffic Engineering approaches attempt to leverage this potential by switching off those parts of the networks that are not required for the routing of specific traffic matrices. When traffic changes, the approaches need to adapt rapidly, which is hard to achieve given the complexity of the problem. We take a fundamentally different approach: instead of considering a specific traffic matrix, we rely on a traffic-oblivious routing scheme. We discuss the NP-hard problem of activating as few connections as possible while still guaranteeing that any down-scaled traffic matrix $\varrho\cdot T$ can be routed, where $\varrho \in (0,1)$ and $T$ is any traffic matrix routable in the original network. We present a $\max(\frac{1}{\varrho\cdot\lambda_{\text{min}}},2)$-approximation algorithm for this problem, with $\lambda_{\text{min}}$ denoting the minimum number of connections between any two connected routers. Additionally, we propose two post-processing heuristics to further improve solution quality. Our evaluation shows that we can quickly generate near-optimal solutions. By design, our method avoids the need for frequent reconfigurations and offers a promising direction to achieve practical energy savings in backbone networks.

cs.NI↗

arXiv2025

Prompting in Practice: Investigating Software Practitioners' Use of Generative AI Tools

The use of generative AI (GenAI) tools has fundamentally transformed software development. Central to this shift is prompt engineering, the practice of crafting textual prompts to guide GenAI tools in generating useful content. Although prompt engineering has emerged as a critical skill, prior research has focused primarily on cataloging of prompting techniques, with limited attention to how software practitioners employ GenAI within real-world development workflows. To address this gap, this study presents a systematic investigation of practitioners' integration of GenAI tools into software development, drawing on a rigorous survey that examines prompting strategies, conversation patterns, and reliability assessments across core software development tasks. We surveyed 72 software practitioners who actively use GenAI to characterize AI usage patterns throughout the development process. By combining qualitative and quantitative analyses of the survey responses, we identified 13 key findings that describe how prompting is performed in practice. Our study shows that while code generation is nearly universal, proficiency strongly correlates with the use of GenAI for more nuanced tasks such as debugging and code review. Practitioners also tend to favor iterative multi-turn conversations to single-shot prompting. Documentation tasks are perceived as most reliable, while complex code generation and debugging remain major challenges. Our findings provide an empirical view of practitioner practices, ranging from basic code generation to deeper integration of GenAI into development workflows, enabling us to offer recommendations for improving both GenAI tools and the ways practitioners interact with them.

cs.SE↗

arXiv2025

Developers' Perspectives on Software Licensing: Current Practices, Challenges, and Tools

Most modern software products incorporate open-source components, requiring development teams to maintain compliance with each component's licenses. Noncompliance can have significant financial, legal, and reputational repercussions. Although some organizations may seek advice from legal practitioners to assist with licensing tasks, developers still play a key role in this process. To this end, it is essential to understand how developers approach licensing compliance tasks, the challenges they encounter, and the tools they use. This work studies these aspects of software licensing practices through a study - conducted by a joint team of software engineering and legal researchers - consisting of a survey with 58 software developers and 7 follow-up interviews. The study resulted in 13 key findings regarding the current state of practice. We discuss the implications of our findings and offer directions for future research, as well as actionable recommendations.

cs.SE↗

arXiv2024

Backdoor attacks on DNN and GBDT -- A Case Study from the insurance domain

Machine learning (ML) will likely play a large role in many processes in the future, also for insurance companies. However, ML models are at risk of being attacked and manipulated. In this work, the robustness of Gradient Boosted Decision Tree (GBDT) models and Deep Neural Networks (DNN) within an insurance context will be evaluated. Therefore, two GBDT models and two DNNs are trained on two different tabular datasets from an insurance context. Past research in this domain mainly used homogenous data and there are comparably few insights regarding heterogenous tabular data. The ML tasks performed on the datasets are claim prediction (regression) and fraud detection (binary classification). For the backdoor attacks different samples containing a specific pattern were crafted and added to the training data. It is shown, that this type of attack can be highly successful, even with a few added samples. The backdoor attacks worked well on the models trained on one dataset but poorly on the models trained on the other. In real-world scenarios the attacker will have to face several obstacles but as attacks can work with very few added samples this risk should be evaluated.

cs.LG↗

arXiv2024

Training Gradient Boosted Decision Trees on Tabular Data Containing Label Noise for Classification Tasks

Label noise, which refers to the mislabeling of instances in a dataset, can significantly impair classifier performance, increase model complexity, and affect feature selection. While most research has concentrated on deep neural networks for image and text data, this study explores the impact of label noise on gradient-boosted decision trees (GBDTs), the leading algorithm for tabular data. This research fills a gap by examining the robustness of GBDTs to label noise, focusing on adapting two noise detection methods from deep learning for use with GBDTs and introducing a new detection method called Gradients. Additionally, we extend a method initially designed for GBDTs to incorporate relabeling. By using diverse datasets such as Covertype and Breast Cancer, we systematically introduce varying levels of label noise and evaluate the effectiveness of early stopping and noise detection methods in maintaining model performance. Our noise detection methods achieve state-of-the-art results, with a noise detection accuracy above 99% on the Adult dataset across all noise levels. This work enhances the understanding of label noise in GBDTs and provides a foundation for future research in noise detection and correction methods.

cs.LG↗

arXiv2023

Green Segment Routing for Improved Sustainability of Backbone Networks

Improving the energy efficiency of Internet Service Provider (ISP) backbone networks is an important objective for ISP operators. In these networks, the overall traffic load throughout the day can vary drastically, resulting in many backbone networks being highly overprovisioned during periods of lower traffic volume. In this paper, we propose a new Segment Routing (SR)-based optimization algorithm that aims at reducing the energy consumption of networks during such low-traffic periods. It uses the traffic steering capabilities of SR to remove traffic from as many links as possible to allow the respective hardware components to be switched off. Furthermore, it simultaneously ensures that solutions comply to additional operator requirements regarding the overall Maximum Link Utilization in the network. Based on data from a Tier-1 ISP and a public available dataset, we show that our approach allows for up to 70 % of the overall linecards to be switched off, corresponding to an around 56% reduction of the overall energy consumption of the network in times of low traffic demands.

cs.NI↗

arXiv2023

Green Traffic Engineering by Line Card Minimization

Green Traffic Engineering encompasses network design and traffic routing strategies that aim at reducing the power consumption of a backbone network. We argue that turning off linecards is the most effective approach to reach this goal. Thus, we investigate the problem of minimizing the number of active line cards in a network while simultaneously allowing a multi-commodity flow being routed and keeping the maximum link utilization below a certain threshold. In addition to proving this problem to be NP-hard, we present an optimal ILP-based algorithm as well as a heuristic based on 2-Segment Routing. Lastly, we evaluate both approaches on real-world networks obtained from the Repetita Framework and a globally operating Internet Service Provider. The results of this evaluation indicate that our heuristic is not only close to optimal but significantly faster than the optimal algorithm, making it viable in practice.

cs.NI↗

arXiv2018

Dynamical structure factor in the non-Abelian phase of the Kitaev honeycomb model in the presence of quenched disorder

Kitaev's model of spins interacting on a honeycomb lattice describes a quantum spin-liquid, where an emergent static $\mathbb{Z}_2$ gauge field is coupled to Majorana fermions. In the presence of an external magnetic field and for a range of interaction strengths, the system behaves as a gapped, non-Abelian quantum spin-liquid. In this phase, the vortex excitations of the emergent $\mathbb{Z}_2$ gauge field have Majorana zero modes bound to them. Motivated by recent experimental progress in measuring and characterizing real materials that could exhibit spin-liquid behavior, we analytically calculate the dynamical spin structure factor in the non-Abelian phase of the Kitaev's honeycomb model. In particular, we treat the case of quenched disorder in the vortex configurations. Our calculations reveal a peak in the low-energy dynamical structure factor that is a signature of the spin-liquid behavior. We map the effective Hamiltonian to that of a chiral p-wave superconductor by using the Jordan-Wigner transformation. Subsequently, we analytically calculate the wave functions of the Majorana zero modes, the energy splitting for finite separation of the vortices and finally, the dynamical structure factor in presence of quenched disorder.

cond-mat.str-el↗

arXiv2017

Korshunov instantons in a superconductor at elevated bias current

Dissipation even at zero temperature reduces quantum fluctuations and tends to localize particles. A notable exception is the nonlinear dissipation due to quasiparticle tunneling in a Josephson junction. It is well known that quasiparticle dissipation does not prevent the phase particle in a superconducting junction to coherently tunnel to the next-nearest minimum even though tunneling to the nearest minimum is suppressed. The reason is that the dissipative action admits an instanton solution, the so-called Korshunov instanton. Here, we analyze this model at elevated bias current $I$. We find that besides the known regime where the logarithm of the tunneling rate scales as $I^{2/3}$ there is novel regime with a scaling $I^2$. We argue that the novel regime opens an opportunity for an experimental verification of the Korshunov instantons as the elevated bias current leads to a larger tunneling rate compared to earlier results.

cond-mat.supr-con↗

arXiv2016

Simulation of supersymmetric quantum mechanics in a Cooper-pair box shunted by a Josephson rhombus

Supersymmetries in quantum mechanics offer a way to obtain degeneracies in the excitation spectrum which do not originate from selection rules. The mechanism behind the degeneracies is the same as the one that leads to the miraculous cancellations of divergences in supersymmetric field theories found in the high energy physics context. Even though of importance, there is up to now no realistic proposal of non-integrable systems that show level degeneracies due to a supersymmetric structure. Here, we propose an implementation of a quantum-mechanical supersymmetry in a Cooper-pair box shunted by a Josephson junction rhombus which is effectively $π$-periodic in the superconducting phase difference. For a characteristic ratio between the strength of the $2π$- and the $π$-periodic junction, we find a two-fold degeneracy of all the energy levels all the way from the weak junction/charge qubit limit to the strong junction/transmon regime. We provide explicit values for the parameters of the rhombus and show that tuning in and out of the supersymmetric point is easily achieved by varying an external gate voltage. We furthermore discuss a microwave experiment to detect the supersymmetry and conclude that it could indeed be simulated with currently existing Josephson junction technology.

cond-mat.mes-hall↗

arXiv2016

Universal power-law decay of electron-electron interactions due to nonlinear screening in a Josephson junction array

Josephson junctions are the most prominent nondissipative and at the same time nonlinear elements in superconducting circuits allowing Cooper pairs to tunnel coherently between two superconductors separated by a tunneling barrier. Due to this, physical systems involving Josephson junctions show highly complex behavior and interesting novel phenomena. Here, we consider an infinite one-dimensional chain of superconducting islands where neighboring islands are coupled by capacitances. We study the effect of Josephson junctions shunting each island to a common ground superconductor. We treat the system in the regime where the Josephson energy exceeds the capacitive coupling between the islands. For the case of two offset charges on two distinct islands, we calculate the interaction energy of these charges mediated by quantum phase slips due to the Josephson nonlinearities. We treat the phase slips in an instanton approximation and map the problem onto a classical partition function of interacting particles. Using the Mayer cluster expansion, we find that the interaction potential of the offset charges decays with an universal inverse-square power law behavior.

cond-mat.mes-hall↗

arXiv2015

Second-order coherence of microwave photons emitted by a quantum point contact

Shot-noise of electrons that are transmitted with probability $T$ through a quantum point contact (biased at a voltage $V_0$) leads to a fluctuating current that in turn emits radiation in the microwave regime. By calculating the Fano factor $F$ for the case where only a single channel contributes to the transport, it has been shown that the radiation produced at finite frequency $ω_0$ close to $e V_0/\hbar$ and at low temperatures is nonclassical with sub-Poissonian statistics ($F<1$). The origin of this effect is the fermionic nature of the electrons producing the radiation, which reduces the probability of simultaneous emission of two or more photons. However, the Fano factor, being a time-averaged quantity, offers only limited information about the system. Here, we calculate the second-order coherence $g^{(2)}(τ)$ for this source of radiation. We show that due to the interference of two contributions, two photon processes (leading to bunching) are completely absent at zero temperature for $T=50\,\%$. At low temperatures, we find a competition of the contribution due to Gaussian current-current fluctuations (leading to bunching) with the one due to non-Gaussian fluctuations (leading to antibunching). At slightly elevated temperatures, the non-Gaussian contribution becomes suppressed whereas the Gaussian contributions remain largely independent of temperature. We show that the competition of the two contributions leads to a nonmonotonic behavior of the second-order coherence as a function of time. As a result, $g^{(2)}(τ)$ obtains a minimal value for times $τ^* \simeq ω_0^{-1}$. Close to this time, the second-order coherence remains below 1 at temperatures where the Fano factor is already above 1. We identify realistic experimental parameters that can be used to test the sub-Poissonian nature of the radiation.

cond-mat.mes-hall↗