Innovation Through Information Systems

Extracting information from invoices is a highly structured, recurrent task in auditing. Automating this task would yield efficiency improvements, while simultaneously improving audit quality. The challenge for this endeavor is to account for the text layout on invoices and the high variety of layouts across different issuers. Recent research has proposed graphs to structurally represent the layout on invoices and to apply graph convolutional networks to extract the information pieces of interest. However, the effectiveness of graph-based approaches has so far been shown only on datasets with a low variety of invoice layouts. In this paper, we introduce a graph-based approach to information extraction from invoices and apply it to a dataset of invoices from multiple vendors. We show that our proposed model extracts the specified key items from a highly diverse set of invoices with a macro $${F}_{1}$$ F 1 score of 0.8753.

Research on digital platform ecosystems is growing rapidly. While the relevance of third-party applications is commonly known, scholars have made only minor attempts to analyze knowledge sharing between platform owners and third-party developers. We find that third-party application development is a knowledge intensive task that requires knowledge to cross organizational boundaries. In this paper, we use computational analytic methods to analyze knowledge sharing in a digital platform ecosystem. We collected trace data about a third-party developer ecosystem with frequent knowledge exchange between the platform owner and third-party developers. We developed a web scraper and retrieved all 4866 pages of SAP’s developer community that were tagged ‘SAP Cloud Platform’. Next, we used text mining to render a topic model. Based on the latent dirichlet allocation algorithm, we extracted 25 topics that were frequently discussed in the community. We clustered the topics into the following six meta-topics: User Accounts and Authentication, Connectivity, Cloud Database, Specific Technologies, SAP Resources, and Installation. Platform owners can use our approach to (1) identify frequently discussed topics, (2) generate meta-knowledge in these topics and (3) use the meta-knowledge to improve their platform core and its boundary resources.

In this paper we elaborate the opportunity of using natural language processing to analyze scientific content both, from a practical as well as a theoretical point of view. Firstly, we conducted a literature review to summarize the status quo of using natural language processing for analyzing scientific content. We could identify different approaches, e.g., with the aim of clustering and tagging publications or to summarize scientific papers. Secondly, we conducted a case study where we used our proposed natural language processing pipeline to analyze scientific content about computer vision available at the database IEEE. Our method helped us to identify emerging trends in the recent years and give an overview of the field of research.

Recommender systems (RS) play a key role in e-commerce by pre-selecting presumably interesting products for customers. Hybrid RSs using a weighted average of individual RSs’ predictions have been widely adopted for improving accuracy and robustness over individual RSs. While for regression tasks, approaches to estimate optimal weighting schemes based on individual RSs’ out-of-sample errors exist, there is scant literature in classification settings. Class prediction is important for RSs in e-commerce, as here item purchases are to be predicted. We propose a method for estimating weighting schemes to combine classifying RSs based on the variance-covariance structures of the errors of individual models’ probability scores. We evaluate the approach on a large real-world e-commerce data set from a European telecommunications provider, where it shows superior accuracy compared to the best individual model as well as a weighting scheme that averages the predictions using equal weights.

Recommender Systems provide users with recommendations for potential items of interest in applications like e-commerce and social media. User information such as past item ratings and personal data can be considered as inputs of these systems. In this study, we aim to utilize a trust-graph-based Neural Network in the recommendation process. The proposed method tries to increase the performance of graph-based RSs by considering the inferred level of trust and its evolution. These recommendations will not only be based on the user information itself but will be fueled by information about associates in the network. To improve the system performance, we develop an attention mechanism to infer a level of trust for each connection in the network. As users are likely to be influenced more by those whom they trust the most, our method might lead to more personalized recommendations, which is likely to increase the user experience and satisfaction.

Although several promising use cases for artificial intelligence (AI) for manufacturing companies have been identified, these are not yet widely used. Existing literature covers a variety of frameworks, methods and processes related to AI systems. However, the application of AI systems in manufacturing companies lacks a uniform understanding of components and functionalities as well as a structured process that supports developers and project managers in planning, implementing, and optimizing AI systems. To close this gap, we develop a generic conceptual model of an AI system for the application in manufacturing systems and a four-phase model to guide developers and project managers through the realization of AI systems.

This paper introduces a framework for managing bias in machine learning (ML) projects. When ML-capabilities are used for decision making, they frequently affect the lives of many people. However, bias can lead to low model performance and misguided business decisions, resulting in fatal financial, social, and reputational impacts. This framework provides an overview of potential biases and corresponding mitigation methods for each phase of the well-established process model CRISP-DM. Eight distinct types of biases and 25 mitigation methods were identified through a literature review and allocated to six phases of the reference model in a synthesized way. Furthermore, some biases are mitigated in different phases as they occur. Our framework helps to create clarity in these multiple relationships, thus assisting project managers in avoiding biased ML-outcomes.

Conversational agents (CAs) have become integral parts of providers’ service offerings, yet their potential is not fully exploited as users’ acceptance and usage of CAs are often limited. Whereas previous research is rather technology-oriented, our study takes a user-centric perspective on the phenomenon. We conduct a systematic literature review to summarize the determinants of individuals’ acceptance, adoption, and usage of CAs that have been examined in extant research, followed by an interview study to identify potential for further research. In particular, five concepts are proposed for further research: personality, risk aversion, cognitive style, self-efficacy, and desire for control. Empirical studies are encouraged to assess the impact of these user-specific concepts on individuals’ decision to use CAs to eventually inform the design of CAs that facilitate users’ acceptance, adoption, and use. This paper intends to contribute to the body of knowledge about the determinants of CA usage.

Voice assistants (VAs) such as Google Assistant and Amazon Alexa are spreading rapidly. They offer users the opportunity to order products online in a spoken dialogue (voice commerce). However, the widespread use of voice commerce is hindered by a lack of satisfaction and trust among VA users. This study investigates whether social cues and the accompanying perception of the VA’s humanness and social presence can overcome existing obstacles in voice commerce. The empirical comparison (N = 323) of two VAs (low vs. high level of social cues) shows that providing VAs with more cues increases user satisfaction. Nevertheless, the analysis does not reveal entirely positive effects on perceived trust and its dimensions of benevolence, competence, and integrity. Surprisingly, users had less trust in the integrity of a VA with more social cues. For a differentiated view, a more in-depth analysis of the individual cues and their interactions is required.

In this paper, artificial neural networks (ANNs) are developed to predict traffic volumes using traffic sensor data from the city of Darmstadt as a basis for future smart mobility solutions. After processing the acquired sensor data, information about the current traffic situation can be derived and events such as rush hour, weekends or holidays can be identified. Based on current research findings in the field of traffic forecasting using neural networks, our work shows the first best practices for modeling the traffic volume and an associated traffic forecast. A Long Short-Term Memory (LSTM) network is shown to be superior to a Deep Neural Network (DNN) in terms of prediction quality and prediction horizon. Furthermore, it is discussed whether the enrichment of the training data with additional time and weather data enables an increase of the forecast accuracy. In the sense of a design-theoretical approach, design requirements and design principles for the development of an ANN in a traffic-specific context are derived.

Chatbot research is currently on its rise since many researchers focus on this topic from different perspectives. Thereby, the focus mostly lies on application areas that originate from business contexts. However, application areas and potential outcomes are already subject to research. The business perspective on influencing factors for an application of chatbots at workplaces or their corresponding challenges is underrepresented as less to none research exists. Therefore, we targeting this research gap by an empirical cross-section interview study with 29 domain experts for the application of chatbots at the digital workplace. We categorize the findings with an extension of the TOE-Framework and show that in the core categories of technological, organizational, individual, and environmental 11 sub-influencing factors exist. Furthermore, we also identify 36 challenges, which are relevant in the particular influencing factors.

Artificial Intelligence (AI) provides organizations with vast opportunities of deploying AI for competitive advantage such as improving processes, and creating new or enriched products and services. However, the failure rate of projects on implementing AI in organizations is still high, and prevents organizations from fully seizing the potential that AI exhibits. To contribute to closing this gap, we seize the unique opportunity to gain insights from five organizational cases. In particular, we empirically investigate how the unique characteristics of AI – i.e. experimental character, context sensitivity, black box character, and learning requirements – induce challenges into project management, and how these challenges are addressed in organizational (socio-technical) contexts. This shall provide researchers with an empirical and conceptual foundation for investigating the cause-effect relationships between the characteristics of AI, project management, and organizational change. Practitioners can benchmark their own practices against the insights to increase the success rates of future AI implementations.

The effective use of artificial intelligence promises significant business value. Effective use, however, requires a thorough exploration of its strengths and weaknesses from different perspectives. Information systems research is particularly invested in the management and use of artificial intelligence in organizations. It has proposed the use of cognitive functions to guide this exploration. In this paper, we evaluate the usefulness of such a cognitive functions lens for a relatively mature application of artificial intelligence, predictive maintenance. Our evaluation is informed by the insights we collected from an embedded single-case study. We find that a cognitive functions lens can indeed be a useful tool to explore artificial intelligence. In particular, it can aid the allocation of tasks between human agents and artificial intelligence-based systems and the design of human-AI hybrids. It is particularly helpful for those who investigate the management of artificial intelligence.

Convolutional neural networks (CNNs) offer great potential for business applications because they enable real-time object recognition. However, their training requires structured data. Crowdsourcing constitutes a popular approach to obtain large databases of manually-labeled images. Yet, the process of labeling objects is a time-consuming and cost-intensive task. In this context, augmented reality provides promising solutions by allowing an end-to-end process of capturing objects, directly labeling them and immediately embedding the data in training processes. Consequently, this paper deals with the development of an object labeling application for crowdsourcing communities following the design science research paradigm. Based on seven issues and twelve corresponding meta-requirements, we developed an AR-based prototype and evaluated it in two evaluation cycles. The evaluation results reveal that the prototype facilitates the process of object detection, labeling and training of CNNs even for inexperienced participants. Thus, our prototype can help crowdsourcing communities to render labeling tasks more efficient.

Labeling training data constitutes the largest bottleneck for machine learning projects. In particular, text classification via machine learning is widely applied and investigated. Hence, companies have to label a decent amount of texts manually in order to build appropriate text classifiers. Obviously, labeling texts manually is associated with time and expenses. Against this background, research started to develop approaches exploiting the knowledge contained in unlabeled texts by learning sophisticated text representations or labeling some of the texts in an automated manner. However, there is still a lack of integrated approaches, considering both types of approaches to further reduce time and expenses for labeling texts. To address this problem, we propose a new hybrid text classification approach combining recent text representations and automated labeling approaches in an integrated perspective. We demonstrate and evaluate our approach using the case of a German bank where the approach could be applied successfully.

Organizations are increasingly adopting artificial intelligence (AI) for business processes. AI-based recommendations aim at supporting users in decision-making, e.g., by pre-filtering options. However, users can often hardly understand how these recommendations are developed. This issue is called “black box problem”. In the context of Human Resources Management, this leads to new questions regarding the acceptance of AI-based recommendations in the recruiting process. Therefore, we develop a model based on the theory of planned behavior explaining the relation between the user’s perception of the black box problem and the attitude toward AI-based recommendations distinguishing between a mandatory and voluntary use context. We conducted 21 interviews with experts from recruiting and AI. Our results show that the perception of the black box problem conceptualized by the awareness and the evaluated relevance relates to the user’s attitude toward AI-based recommendations. Further, we show that the use context has a moderating effect on that relation.

Phishing attacks are the primary cause of data and security breaches in businesses, public institutions, and private life. Due to inherent limitations and users’ high susceptibility to increasingly sophisticated phishing attempts, existing anti-phishing measures cannot realize their full potential. Against this background, we utilize methods from the emerging research field of Explainable Artificial Intelligence (XAI) for the design of a user-focused anti-phishing measure. By leveraging the power of state-of-the-art phishing detectors, our approach uncovers the words and phrases in an e-mail most relevant for identifying phishing attempts. We empirically show that our approach reliably extracts segments of text considered relevant for the discrimination between genuine and phishing e-mails. Our work opens up novel prospects for phishing prevention and demonstrates the tremendous potential of XAI methods beyond applications in AI.

By combining recent advances in Natural Language Processing and Conversational Agent (CAs), we suggest a new form of human-computer interaction for individuals to receive formative feedback on their argumentation to help them to foster their logical reasoning skills. Hence, we introduce ArgueBot, a conversational agent, that provides adaptive feedback on students’ logical argumentation. We, therefore, 1) leveraged a corpus of argumentative student-written peer-reviews in German, 2) trained, tuned, and benchmarked a model that identifies claims, premises and non-argumentative sections of a given text, and 3) built a conversational feedback tool. We evaluated ArgueBot in a proof-of-concept evaluation with students. The evaluation results regarding technology acceptance, the performance of our trained model, and the qualitative feedback indicate the potential of leveraging recent advances in Natural Language Processing for new human-computer interaction use cases for scalable educational feedback.

Conversational agents (CAs) are rapidly changing the way humans and computers interact. Through developments in natural language processing, CAs are increasingly capable of conducting human-like conversations with users. Furthermore, human-like features (e.g., having a name or an avatar) lead to positive user reactions as if they were interacting with a real human conversational partner.CAs promise to replace or supplement traditional interactions between humans (e.g., counseling, interviews). One field of CA-human interaction that is not yet fully understood in developing human-like CAs is donating to a good cause. Notably, many charities rely on approaching people on the streets to raise funds.Against this background, the questions arise: How should a CA for raising funds for non-profit organizations be designed and how does human-like design of a CA influence the user’s donation behavior. To explore these two questions, we conducted a 2 × 2 experiment with 134 participants.

In the human interaction with CAs, research has shown that elements of persuasive system design, such as praise, are perceived differently when compared to traditional graphical interfaces.In this experimental study, we will extend our knowledge regarding the relation of persuasiveness (namely dialog support), anthropomorphically designed CAs, and task performance. Within a three-conditions-between-subjects design, two instances of the CA are applied within an online experiment with 120 participants. Our results show that anthropomorphically designed CAs increase perceived dialog support and performance but adding persuasive design elements can be counterproductive. Thus, the results are embedded in the discourse of CA design for task support.

Driven by technological advances, smart personal assistants (SPA) have gained importance in human–computer interaction. SPA can influence user behavior and persuade users to reach a specific outcome. However, users often lack the motivation to interact with SPA. One way to support this interaction is persuasive system design – considering concepts as gamification and nudging. Although SPA research has increased recently, there is still no shared knowledge about persuasive designs. Therefore, we aim to identify the current state-of-the-art Design for persuasive SPA to understand how interactions and designs can be improved. Thus, we conduct a systematic literature analysis to represent how gamification and digital nudging are used to design SPA and conclude if and how those concepts can support SPA interactions. Consequently, we contribute to theory, providing better understanding about SPA interaction and design to make SPA more engaging and entertaining. Practitioners can use this contribution for persuasive SPA designs.

While regional cues are omnipresent in offline consumer life, the use of regional imagery is a still emerging trend in online retail. Applying a multi-method approach, we investigate the use of regional imagery on web interfaces and its effects on consumer behavior. We find that social, nature, and regional imagery is frequently used on energy provider websites and identify cityscapes and monuments as primary regional motives. Further, we outline an experiment to assess whether regional imagery promotes trust within online retail and how the presence of regionality interacts with the concepts of Social Presence and Nature Presence. Our contribution is twofold: First, we propose the psychological construct of Regional Presence to the IS literature, link it to theory, and describe its application in online retail. Second, we sketch out an experimental design to systematically study the effects of regional imagery on web interfaces.

Enabled by technological advancements, a contemporary form of technology use that particularly became popular are online multiplayer video games, which are played with others in real time. Besides various positive impacts on the user experience (e.g., fun, additional social exchange) adverse consequences have occurred as well (e.g., stress, anger). Most recently, a sincere problem gaining increased attention is toxic behavior (i.e., a behavior spreading negative effects and bad mood during play). With our study, we propose a way to handle toxic behavior on a level of video game design by using a multi-method approach. First, we will consult the online disinhibition effect and its antecedents to identify design related relationships. Afterwards, we will conduct a qualitative workshop engaging video game designers and players to reshape in-game experiences by incentivizing players to buffer toxicity.

This paper analyzes data protection challenges and possible solutions associated with the usage of the blockchain (BC) technology from the perspective of 94 German companies and organizations. This paper clusters 537 data protection-relevant statements into three subject areas: (1) relevance of data protection in BC, (2) articulated challenges and (3) proposed solutions. Each group is then collated with insights from computer science. The results show that a majority of the respondents do see data protection issues with using BC, which mainly relate to data erasure and identifying the data controller. However, the majority also consider these problems to be solvable utilizing already available technologies, e.g. off-chain storage, encryption, pseudonymization or usage of private BCs. Comparing these proposals with the findings in computer science literature shows that especially off-chain storage, encryption and redactable blockchains can be regarded as adequate solutions.

Awareness of cybersecurity topics facilitates software developers to produce secure code. This awareness is especially important in industrial environments for the products and services in critical infrastructures. In this work, we address how to raise awareness of software developers on the topic of secure coding. We propose the “CyberSecurity Challenges”, a serious game designed to be used in an industrial environment and address software developers’ needs. Our work distills the experience gained in conducting these CyberSecurity Challenges in an industrial setting. The main contributions are the design of the CyberSecurity Challenges events, the analysis of the perceived benefits, and practical advice for practitioners who wish to design or refine these games.

The relevance of adequate privacy and security behavior in the digital realm is higher than ever. However, the exact relationship between privacy and security behavior is rarely discussed in the literature. This study investigates this relationship and the role of socio-demographic factors (gender, age, education, political ideology) in such behavior. Results of a survey among German private users (N = 1,219) show that privacy and security behavior are only weakly correlated and not similarly influenced by socio-demographic factors. While security behavior significantly differs between age and education groups (younger and less educated show less security behavior), no such differences exist for privacy behavior. Additionally, political ideology has no influence on privacy and security behavior. Thus, this study sheds light on the concepts of privacy, security and corresponding behavior and emphasizes the need for a fine-grained differentiation if either privacy or security behavior is to be improved.

Higher legal standards with regards to data protection of individuals such as the European General Data Protection Regulation (GDPR) increase the pressure on developing lawful technologies. The development requires feedback from stakeholders such as legal experts that lack technical knowledge but are required to understand IT artifacts. As a solution, patterns can support interdisciplinary system development. We demonstrate how design patterns can support legal experts in arguing about technologies in court by introducing a law simulation study which is a well-known evaluation method in law. Our results show that patterns support legal experts in their argumentation about technologies in court. We provide theoretical contributions concerning cognitive fit theory about how patterns act as a bridge between the internal and external representation of problems and improve problem-solving performance related to the legal assessment of technology. In addition, we provide practical guidance for codifying and communicating design knowledge through patterns.

Future used car markets may use personal data to reduce information asymmetries between car sellers and buyers, e. g. on past driving behavior. Reducing information asymmetries is attractive for used car platforms as they can move from pure information provision to orchestrating transactions. However, car sellers and buyers have to agree to sharing personal data. What kind of data is interesting for them? Under what circumstances are they willing to share this data? What should a platform do to support data sharing? We explore those research questions as part of the Cardossier project by conducting experiments with the Car-Market Game, simulating a future car market. The results indicate that there is no market for pure personal data (e. g. photographs of sellers), but there is a market for car usage data. From future used car platforms the participants expect disclosure control and disclosure transparency in an environment free of interpersonal trust.

This paper gives an overview of current research areas considering GDPR and blockchain. It is shown that GDPR is often seen as a problem, limiting blockchain use cases. However, approaches towards more data protection for the data subjects based on blockchain technology emerge. In this paper, we evaluate a first step towards a GDPR enforcing blockchain by using a combination of smart contracts within Hyperledger Fabric, evaluating if a joint controllership agreement is in place. Such agreement is required for joint controller to process personal data. Based on this rather simple use case evaluation, it is discussed that a combination of the different research areas around GDPR and blockchain should be further evaluated and combined, aiming to GDPR enforcing blockchain systems.

Social media have become part of millions of users’ everyday life, leading to the proliferation of the daily stressors associated with them, particularly social media-induced overloads. Therefore, understanding the individual characteristics that enable users to resist such stress factors and ultimately buffer negative follow-up effects, such as exhaustion and discontinuance behavior, is important for researchers and practitioners. Grounded in psychological resilience theory, we examine if a user’s resilience (one’s ability to bounce back) has the power to mitigate the effects of this critical chain of influence by inhibiting the stressors. Structural equation modelling on survey data from 194 social network users confirms that resilience decreases perceived information and social overload. We also find that self-efficacy is a protective factor leading to resilience. Therein, this short paper raises awareness on resilience’s function as a shield against the adverse effects of social media and provides a comprehensive outlook for future research.

The emergence of digital work leads to an increasing number of teams that collaborate virtually. The physical absence of team members and other problems that occur due to the virtual context hinder the formation of team identification, which however is essential for employee’s work motivation and team effectiveness. Our study addresses the question of how companies can improve team identification in virtual teams. Combining a literature review with interviews with five case teams, we derive an overview of solutions to improve team identification and link these to problems in virtual team identification. Our results provide guidance to practitioners and we further derive propositions for future research based on identified research gaps and inconsistencies.

Conventional methods for tie prediction analysis in social media networks are often code-intensive and encompass complex steps. Against this backdrop, we used design science research to develop a no-code tie prediction analysis platform. Our evaluation indicates that the platform significantly reduces tie prediction analysis complexity and, depending on the network size, also total prediction time. Moreover, it maintains a prediction accuracy similar to that of conventional, code-intensive methods. Thus, our artifact substantially facilitates tie prediction analysis for social media network researchers and practitioners.

The ongoing digital transformation shapes the world of information discovery and dissemination for investment decisions. Social investment platforms offer the possibility for non-professionals to publish financial analyst reports on company development and earnings forecast and give investment recommendations similar to those provided by traditional sell-side analysts. This phenomenon of “crowd analyst reports” has been found to provide an adequate alternative for non-professional investors. In this study, we examine the informational value of these crowd analyst reports regarding their timeliness in publishing and their originality as for content and opinion. Our findings suggest that crowd analysts strongly rely on previously published institutional reports. Therefore, crowd analysts do not pose a threat to institutional analysts at this time, however, they provide a more accessible information basis and improve decision-making for individual investors.

The workforce of an organization plays a critical role for the success (or failure) of digital innovation; they need to have specific skills, which are required for creating the needed digital change. Therefore, organizations need to continuously upskill their workforce. Different ways to prepare and upskill the workforce for the digital future exist. However, a structured approach to guide organizations on how to retrain and upskill their workforce is lacking. In the light of this context, the research goal is to provide an action-oriented guideline in form of design principles supporting organizations to handle digital upskilling. To achieve this goal a hermeneutic literature review and semi-structured expert interviews as well as a focus group discussion have been performed to deduce design principles. Based on an applicability check proposed by Rosemann and Vessey [1] the usefulness and applicability of the resulting 15 design principles in organizations are validated.

Design knowledge on agile role development is still nascent. Following the call for a theoretical and empirical investigation into the formation of roles in agile transformations, we elicited three design goals, six design principles and defined the pathway to achieving these goals. Our concept provides new insights into the dynamic and cohesive character of agile roles and adds role development as a key activity to the core of agile transformation. The results are based on a qualitative research methodology and design science research. We contribute to theory by providing a grounded approach for the situational development of dynamic agile roles based on design goals and principles, while practice can profit from the approach as our concept provides greater flexibility than the strict role definitions of agile frameworks.

Increased digitalization offers today’s organizations novel opportunities to enhance value propositions for customers, but also poses significant challenges for traditional businesses. To navigate through the difficult process of digital transformation in this turbulent environment, organizations need to integrate successful innovation management practices and build organizational resilience.In this paper, we propose a conceptual framework that bridges between these two constructs: We describe innovation management as the continuous activity of anticipating and responding to ongoing trends in an organization’s environment through innovation, whereas we understand organizational resilience as the capability to adapt or transform an organization’s business. By analyzing two illustrative cases, we find indications that a successful digital transformation is not possible without one or the other. Furthermore, we contribute key factors for building organizational resilience and showcase two examples of how to leverage organizational resilience by transforming business models through digital innovation and, thus, avoiding the innovator’s dilemma.

This study is a methodological replication of the PROFUND method for the implementation of digital transformation projects in small and medium-sized enterprises (SMEs). In the original study, a method for improving the Value Co-Creation (VCC) of an SME network was applied in the digital transformation of a network of textile manufacturers to validate four existing propositions for improving the VCC. This study replicates (1) the procedure in the case study, (2) the guidelines for improving the VCC, and (3) the evaluation of the results via the DART model and transfers them to a network in the metal industry. The method was transferable and the propositions could be reconfirmed. The results of this study confirm the results of the original study and thus validate the procedural structure used there.

The IT industry is getting more and more agile. Therefore, it is important to know about the required skills for agile workers in this industry. In this study, we analyzed 1000 job advertisements from online job portals to determine the differences in skill requirements for agile workers in the IT industry in Germany versus the U.S. We found that searches for non-technical skills are greater in the U.S. than in Germany. Test and Requirements Management are the most important management concept searches in both countries. JavaScript is searched more often in the U.S. This study contributes to a better understanding of the required skills of agile workers in Germany and the U.S.

Digital B2B platforms are becoming increasingly important for value co-creation in today’s business networks, leading to the emergence of a diverse landscape of platforms and intensifying research efforts. Yet, practitioners and researchers alike lack a means to structure existing knowledge and distinguish between different B2B platforms. In this paper, we apply Nickerson et al.’s method for taxonomy development to derive a taxonomy of B2B co-creation platforms drawing on 36 research articles and 63 real-world platform cases. We find 17 dimensions that describe B2B co-creation platforms in terms of their platform architecture, their actor ecosystem, and their value creation process. Thereby, we contribute to research and practice: First, we provide a holistic perspective on B2B co-creation platforms by aggregating existing knowledge and identifying the fundamental properties relevant for their distinction. Second, we provide a decision aid for practitioners to evaluate which platform to join or how to design B2B co-creation platforms.

In the building supply industry, there is increasing price pressure for semi-finished products. This price pressure is strongly influenced by the commoditisation of the product business. One way to differentiate from competitors in product-heavy industries such as the construction industry is to develop and offer digital services. Digital services can be used to expand existing product solutions and provide them with additional functionality, as well as to develop completely independent solutions. Problematic for existing supplier companies is the strongly product-focused know-how and the low level of competence in the development of digital solutions. This paper presents the conception and development of a decision support system for the technical implementation of modular services. Therefore, a systematic procedure for requirements specification as well as a user-centred description of the essential functional features is explained. Finally, the system is tested within the framework of an evaluation with 16 test persons from product development and business development in practical case studies.

Platforms have disrupted several business sectors and daily life in general. Platforms facilitate collaboration between different partners, which leads to the emergence of an ecosystem. During recent years, both research fields platforms and ecosystems have made significant progress. Since the terminologies originate from different backgrounds and are put into play in various sectors, a certain vagueness surrounds platforms and ecosystems. The present paper, therefore, adds to academia by providing an ontology – an abstraction of a real-world phenomenon – for platform ecosystems. The ontology comprises concepts from the platforms, business ecosystems, and platform ecosystems domains. The evaluation with three real-world platform ecosystems from different industries verifies that the platform-ecosystem-specific requirements were met in the ontology.

Balancing parking space capacities and distributing capacity information play an important role in modern metropolitan life and urban land use management. They promise not only optimal urban land use and reductions of search time for suitable parking, but also contribute to a lower fuel consumption. Based on a design science research approach we develop a solution to parking space management through deep learning and aspire to design a camera-based, low-cost, scalable, real-time detection of occupied parking spaces. We evaluate the solution by building a prototype to track cars on parking lots that improves prior work by using a TensorFlow deep neural network with YOLOv4 and DeepSORT. Additionally, we design a web interface to visualize parking capacity and provide further information, such as average parking times. This work contributes to camera-based parking space management on public, open-air parking lots.

Mobility is an essential need that requires novel opportunities enabling us to travel more sustainably. In attempting to address this, our university—the University Hildesheim—, located within a city of about 100.000 residents seeks to improve especially the student’s and employee’s arrival approaches to, for instance, reduce greenhouse emissions caused by traffic, relax the current parking situation, and limit traffic jams. By drawing on a literature review, an analysis of a university-wide mobility survey, and several interviews, this study (1) deduced a set of eight requirements for choosing more environmentally-friendly mobility options and (2) developed a mobile application (app) that promotes carpooling through the help of digital nudges. With this, we hope to contribute to current mobility challenges especially due to increased traffic.

The effectiveness of public warning systems (PWS) can be challenged by federal structures as the failure of the first nationwide German “Warntag” (Warning Day) showed. By designing PWS to address specific challenges of federal systems, the effectiveness of public warning might be improved. In this paper, we derive design principles for PWS which aim to address these specific challenges. Based on a thorough literature review, challenges regarding responsibility, coordination, and interoperability, as well as functional and technical requirements for PWS in federal systems were identified. By applying a design-oriented research approach, 16 design principles in the categories strategy and governance, standards and templates, and technology are articulated. The research provides guidance for responsible authorities in federal systems for the implementation or evaluation of public warning systems.

To survive in today's economy, it is no longer enough to stick to old habits and corporate structures. In the age of digitalization, it is much more important to recognize innovation and optimization potential in good time and to constantly question them. In the long term, this can only be ensured by establishing suitable data governance in the company. In addition to data governance, IT, information and corporate governance play an important role in installing organizational frame conditions in companies. In relation to the starting position described above, the current state of scientific research on data governance and the mutual dependencies on the adjacent types of governance: corporate, information and IT governance, will be presented by means of a literature search according to Fettke (2006) and Webster and Watson (2002).

The Internet of Things (IoT) connects humans and machines by means of intelligent technology such as sensors or actuators. This makes it possible to network everyday objects or machines in an industrial environment via the internet. The data generated in this way is also known as Big Data. Master data management (MDM) can offer great potential in dealing with data and data quality by providing a set of guidelines for data management and thus enabling a common view of it. In this paper, different approaches for the use of master data management in the context of IoT are analysed. For this purpose, a classification of the possible uses in the different design or functional areas is given in order to highlight areas of master data management with particular potential for use. The analysed results show that of the three design areas of enterprise-wide MDM, the system level is most frequently represented.

The digitization of businesses provides huge amounts of data that can be leveraged by modern Business Process Management methods. Predictive Business Process Monitoring (PBPM) represents techniques which deal with real-time analysis of currently running process instances and also with the prediction of their future behavior. While many different prediction techniques have been developed, most of the early techniques base their predictions solely on the control­fow characteristic of a business process. More recently, researchers attempt to incorporate additional process-related information, also known as the process context, into their predictive models. In 2018, Di Francescomarino et al. published a framework of existing prediction techniques. Since the young field has evolved greatly since then and context information continue to play a greater role in predictive techniques, this paper describes the process and outcome of updating and extending the framework to include process context dimensions by replicating the literature review of the initial authors.

Process digitalization in health care systems can help to increase treatment quality and foster cost-efficiency at the same time. Several studies have already shown how digitalization can change and automate processes, but do not address the specific needs of dentists. This paper investigates the status quo and the possibilities of process digitalization for dental practices and their interfaces. Based on the assessment of 101 participants of an empirical study and semi-structured interviews with four dentists and four professionals of statutory health insurance we provide insights into the current situation. The analysis reveals that most organizations are in an early stage of digital transformation but are actively working on enhancing the digitalization of their processes. To meet this need, our results deliver a detailed description of prevailing challenges and starting points for process digitalization and optimization in the field of dentistry.

Streaming Video-on-demand (SVOD) services are getting increasingly popular. Current research, however, lacks knowledge about consumers’ content decision processes and their respective influencing factors. Thus, the work reported on in this paper explores socio-technical interrelations of factors impacting content choices in SVOD, examining the social factors WOM, eWOM and peer mediation, as well as the technological influence of recommender systems. A research model based on the Theory of Reasoned Action and the Technology Acceptance Model was created and tested by an n = 186 study sample. Results show that the quality of a recommender system and not the social mapping functionality is the strongest influencing factor on its perceived usefulness. The influence of the recommender system and the influence of the social factors on the behavioral intention to watch certain content is nearly the same. The strongest social influencing factor was found to be peer mediation.

Bitcoin is a cryptocurrency and is considered a high-risk asset class whose price changes are difficult to predict. Current research focusses on daily price movements with a limited number of predictors. The paper at hand aims at identifying measurable indicators for Bitcoin price movements and the development of a suitable forecasting model for hourly changes. The paper provides three research contributions. First, a set of significant indicators for predicting the Bitcoin price is identified. Second, the results of a trained Long Short-term Memory (LSTM) neural network that predicts price changes on an hourly basis is presented and compared with other algorithms. Third, the results foster discussions of the applicability of neural nets for stock price predictions. In total, 47 input features for a period of over 10 months could be retrieved to train a neural net that predicts the Bitcoin price movements with an error rate of 3.52%.