Refine
Document Type
- Article (6)
- Doctoral Thesis (3)
- Part of a Book (2)
- Report (2)
- Conference Proceeding (1)
- Postprint (1)
Is part of the Bibliography
- yes (15)
Keywords
- artificial intelligence (15) (remove)
Institute
- Fachgruppe Betriebswirtschaftslehre (7)
- Hasso-Plattner-Institut für Digital Engineering GmbH (2)
- Institut für Biochemie und Biologie (2)
- Digital Engineering Fakultät (1)
- Extern (1)
- Hasso-Plattner-Institut für Digital Engineering gGmbH (1)
- Institut für Informatik und Computational Science (1)
- Institut für Physik und Astronomie (1)
- Zentrum für Lehrerbildung und Bildungsforschung (ZeLB) (1)
- Öffentliches Recht (1)
How do social changes, new technologies or new management trends affect communication work? A team of researchers at Leipzig University and the University of Potsdam (Germany) observed new developments in related disciplines. As a result, the five most important trends for corporate communications are identified annually and published in the Communications Trend Radar. Thus, Communications managers can identify challenges and opportunities at an early stage, take a position, address issues and make decisions. For 2023, the Communications Trend Radar identifies five key trends for corporate communications: State Revival, Scarcity Management, Unimagination, Parallel Worlds, Augemented Workflows.
Modern datasets often exhibit diverse, feature-rich, unstructured data, and they are massive in size. This is the case of social networks, human genome, and e-commerce databases. As Artificial Intelligence (AI) systems are increasingly used to detect pattern in data and predict future outcome, there are growing concerns on their ability to process large amounts of data. Motivated by these concerns, we study the problem of designing AI systems that are scalable to very large and heterogeneous data-sets.
Many AI systems require to solve combinatorial optimization problems in their course of action. These optimization problems are typically NP-hard, and they may exhibit additional side constraints. However, the underlying objective functions often exhibit additional properties. These properties can be exploited to design suitable optimization algorithms. One of these properties is the well-studied notion of submodularity, which captures diminishing returns. Submodularity is often found in real-world applications. Furthermore, many relevant applications exhibit generalizations of this property.
In this thesis, we propose new scalable optimization algorithms for combinatorial problems with diminishing returns. Specifically, we focus on three problems, the Maximum Entropy Sampling problem, Video Summarization, and Feature Selection. For each problem, we propose new algorithms that work at scale. These algorithms are based on a variety of techniques, such as forward step-wise selection and adaptive sampling. Our proposed algorithms yield strong approximation guarantees, and the perform well experimentally.
We first study the Maximum Entropy Sampling problem. This problem consists of selecting a subset of random variables from a larger set, that maximize the entropy. By using diminishing return properties, we develop a simple forward step-wise selection optimization algorithm for this problem. Then, we study the problem of selecting a subset of frames, that represent a given video. Again, this problem corresponds to a submodular maximization problem. We provide a new adaptive sampling algorithm for this problem, suitable to handle the complex side constraints imposed by the application. We conclude by studying Feature Selection. In this case, the underlying objective functions generalize the notion of submodularity. We provide a new adaptive sequencing algorithm for this problem, based on the Orthogonal Matching Pursuit paradigm.
Overall, we study practically relevant combinatorial problems, and we propose new algorithms to solve them. We demonstrate that these algorithms are suitable to handle massive datasets. However, our analysis is not problem-specific, and our results can be applied to other domains, if diminishing return properties hold. We hope that the flexibility of our framework inspires further research into scalability in AI.
Für die Entwicklung professioneller Handlungskompetenzen angehender Lehrkräfte stellt die Unterrichtsreflexion ein wichtiges Instrument dar, um Theoriewissen und Praxiserfahrungen in Beziehung zu setzen. Die Auswertung von Unterrichtsreflexionen und eine entsprechende Rückmeldung stellt Forschende und Dozierende allerdings vor praktische wie theoretische Herausforderungen. Im Kontext der Forschung zu Künstlicher Intelligenz (KI) entwickelte Methoden bieten hier neue Potenziale. Der Beitrag stellt überblicksartig zwei Teilstudien vor, die mit Hilfe von KI-Methoden wie dem maschinellen Lernen untersuchen, inwieweit eine Auswertung von Unterrichtsreflexionen angehender Physiklehrkräfte auf Basis eines theoretisch abgeleiteten Reflexionsmodells und die automatisierte Rückmeldung hierzu möglich sind. Dabei wurden unterschiedliche Ansätze des maschinellen Lernens verwendet, um modellbasierte Klassifikation und Exploration von Themen in Unterrichtsreflexionen umzusetzen. Die Genauigkeit der Ergebnisse wurde vor allem durch sog. Große Sprachmodelle gesteigert, die auch den Transfer auf andere Standorte und Fächer ermöglichen. Für die fachdidaktische Forschung bedeuten sie jedoch wiederum neue Herausforderungen, wie etwa systematische Verzerrungen und Intransparenz von Entscheidungen. Dennoch empfehlen wir, die Potenziale der KI-basierten Methoden gründlicher zu erforschen und konsequent in der Praxis (etwa in Form von Webanwendungen) zu implementieren.
Artificial intelligence (AI)-based technologies can increasingly perform knowledge work tasks, such as medical diagnosis. Thereby, it is expected that humans will not be replaced by AI but work closely with AI-based technology (“augmentation”). Augmentation has ethical implications for humans (e.g., impact on autonomy, opportunities to flourish through work), thus, developers and managers of AI-based technology have a responsibility to anticipate and mitigate risks to human workers. However, doing so can be difficult as AI encompasses a wide range of technologies, some of which enable fundamentally new forms of interaction. In this research-in-progress paper, we propose the development of a taxonomy to categorize unique characteristics of AI-based technology that influence the interaction and have ethical implications for human workers. The completed taxonomy will support researchers in forming cumulative knowledge on the ethical implications of augmentation and assist practitioners in the ethical design and management of AI-based technology in knowledge work.
Does AI control or support?
(2022)
Many companies are currently investing in artificial intelligence (AI) because of its potential to increase customer satisfaction or financial performance. However, the success rates in implementing AI systems are low, partly due to technology-centric approaches that neglect work practices. This study draws on Bourdieu’s theory of practice to highlight the potential power shift related to AI in customer relationship management, based on the concepts field, capital, and habitus. Two longitudinal case studies were conducted to understand the power shift related to AI implementation. These two AI systems were designed with the objective to support employees. However, subsequently, their implementation changed the balance of power with a significant shift towards more management control, resulting in a devaluation of employees’ work practices. The paper discusses implications for theory and practice in terms of the discrepancies and power shifts following the introduction of AI systems to support customer relationship management.
We and AI
(2021)
Digitizing grocery retailing
(2021)
Multiple emerging technologies both threaten grocers and offer them attractive opportunities to enhance their value propositions, improve processes, reduce costs, and therefore generate competitive advantages. Among the variety of technological innovations and considering the scarcity of resources, it is unclear which technologies to focus on and where to implement them in the value chain. To develop the most probable technology forecast that addresses the application of emerging technologies in the grocery value chain within the current decade, we conduct a two-stage Delphi study. Our results suggest a high relevance of almost all technologies. The panel is only skeptical about three specific projections. As a consequence, grocers are advised to build up knowledge regarding the application of these technologies in the most promising areas of their value chain.
As AI technology is increasingly used in production systems, different approaches have emerged from highly decentralized small-scale AI at the edge level to centralized, cloud-based services used for higher-order optimizations. Each direction has disadvantages ranging from the lack of computational power at the edge level to the reliance on stable network connections with the centralized approach. Thus, a hybrid approach with centralized and decentralized components that possess specific abilities and interact is preferred. However, the distribution of AI capabilities leads to problems in self-adapting learning systems, as knowledgebases can diverge when no central coordination is present. Edge components will specialize in distinctive patterns (overlearn), which hampers their adaptability for different cases. Therefore, this paper aims to present a concept for a distributed interchangeable knowledge base in CPPS. The approach is based on various AI components and concepts for each participating node. A service-oriented infrastructure allows a decentralized, loosely coupled architecture of the CPPS. By exchanging knowledge bases between nodes, the overall system should become more adaptive, as each node can “forget” their present specialization.
On 21 April 2021, the European Commission presented its long-awaited proposal for a Regulation “laying down harmonized rules on Artificial Intelligence”, the so-called “Artificial Intelligence Act” (AIA). This article takes a critical look at the proposed regulation. After an introduction (1), the paper analyzes the unclear preemptive effect of the AIA and EU competences (2), the scope of application (3), the prohibited uses of Artificial Intelligence (AI) (4), the provisions on high-risk AI systems (5), the obligations of providers and users (6), the requirements for AI systems with limited risks (7), the enforcement system (8), the relationship of the AIA with the existing legal framework (9), and the regulatory gaps (10). The last section draws some final conclusions (11).
Research synthesis on simple yet general hypotheses and ideas is challenging in scientific disciplines studying highly context-dependent systems such as medical, social, and biological sciences. This study shows that machine learning, equation-free statistical modeling of artificial intelligence, is a promising synthesis tool for discovering novel patterns and the source of controversy in a general hypothesis. We apply a decision tree algorithm, assuming that evidence from various contexts can be adequately integrated in a hierarchically nested structure. As a case study, we analyzed 163 articles that studied a prominent hypothesis in invasion biology, the enemy release hypothesis. We explored if any of the nine attributes that classify each study can differentiate conclusions as classification problem. Results corroborated that machine learning can be useful for research synthesis, as the algorithm could detect patterns that had been already focused in previous narrative reviews. Compared with the previous synthesis study that assessed the same evidence collection based on experts' judgement, the algorithm has newly proposed that the studies focusing on Asian regions mostly supported the hypothesis, suggesting that more detailed investigations in these regions can enhance our understanding of the hypothesis. We suggest that machine learning algorithms can be a promising synthesis tool especially where studies (a) reformulate a general hypothesis from different perspectives, (b) use different methods or variables, or (c) report insufficient information for conducting meta-analyses.