Business Information Systems

Within the growing amount of data through new applications, processes and technologies in companies, legal frameworks according to the processing of data become more important. The new General Data Protection Regulation (GDPR) especially has the intention, to strengthen the rights of Data Subjects in transparency (e.g. Art. 12) and self-control (e.g. Art 15–22). This research aims to develop non-functional-requirements (NFR) for a Data Transparency System for the category legal-contractual. Therefore, we follow the requirement engineering process according to Rupp [29]. As a general source for the development, qualitative expert interviews have been carried out. In order to extend our findings and form categories, we also did a systematic literature review and a structured text analysis of the GDPR. In total, we were able to generate 18 NFR and organized them into the categories Purpose, Obligation, Ownership, Procedures and Integrity and Transparency.

In the present era of pervasive mobile technologies, interconnecting innovations are increasingly prevalent in our lives. In this evolutionary process, mobile and social media communication systems serve as a backbone for human interactions. When assessing privacy risks related to this, privacy scoring models (PSM) can help quantifying the personal information risks. This paper uses the mobile phone number itself as a basis for privacy scoring. We tested 1,000 random phone numbers for their matching to social media accounts. The results raise concerns how network and communication layers are predominately connected. PSMs will support future organizational sensitivity for data linkability.

REM sleep behavior disorder (RBD) is commonly associated with Parkinson’s disease. In order to find adequate therapy for affected persons and to seek suitable early Parkinson Patterns, the investigation of this phenomenon is highly relevant. The analysis of sleep is currently done by manual analysis of polysomnography (PSG), which leads to divergent scoring results by different experts. Automated sleep stage detection can help deliver accurate, reproducible scoring results. In this paper, we evaluate different machine learning models from the PSG signals for automatic sleep stage detection. The highest accuracy of 87.57% was achieved by using the Random Forest algorithm.

Although the term of big data and related technologies received lots of attention in recent years, many projects are less successful than anticipated. One of the most crucial steps in the planning of a system includes the modeling of the underlying architecture. However, as of now, no standardized approach exists that facilitates the modeling of big data system architectures (BDSA). In this research, a systematic approach is presented that delivers a foundation towards a standard for the modeling of BDSA. Further, a prototype is introduced that automatizes the creation of those models reducing the required effort and simultaneously increasing the maintainability.

In recent years, the amount of data is growing extensively. In companies, spreadsheets are one common approach to conduct data processing and statistical analysis. However, especially when working with massive amounts of data, spreadsheet applications have their limitations. To cope with this issue, we introduce a human-in-the-loop approach for scalable data preprocessing using sampling. In contrast to state-of-the-art approaches, we also consider conflict resolution and recommendations based on data not contained in the sample itself. We implemented a fully functional prototype and conducted a user study with 12 participants. We show that our approach delivers a significantly higher error correction than comparable approaches which only consider the sample dataset.

The discipline of process mining deals with analyzing execution data of operational processes, extracting models from event data, checking the conformance between event data and normative models, and enhancing all aspects of processes. Recently, new techniques have been developed to analyze event data containing uncertainty; these techniques strongly rely on representing uncertain event data through graph-based models capturing uncertainty. In this paper we present a novel approach to efficiently compute a graph representation of the behavior contained in an uncertain process trace. We present our new algorithm, analyze its time complexity, and report experimental results showing order-of-magnitude performance improvements for behavior graph construction.

Personal customer care is one of the advantages of physical retail over its online competition, but cost pressure forces retailers to deploy staff as efficiently as possible resulting in a trend of staff reduction. For staff and managers it becomes harder to keep track of what is happening in a store. Situations that would benefit from intervention like cases of aimless customers, lost children or shoplifting go unnoticed. To this end, real-time tracking systems can provide managers with live data on the current in-store situation, but analysis methods are necessary to actually interpret these data. In particular, anomaly detection can highlight unusual situations that require a closer look. Unfortunately, existing algorithms are not well-suited for a retail scenario as they were designed for different use cases or are slow to compute. To resolve this, we investigate the use of long short-term memory autoencoders, which have recently shown to be successful in related scenarios, for real-time detection of unusual customer behavior. As we demonstrate, autoencoders reconcile the precision of reliable methods that have poor performance with a speed suitable for practical use.

Nowadays, chatbots have become more and more prominent in various domains. Nevertheless, designing a versatile chatbot, giving reasonable answers, is a challenging task. Thereby, the major drawback of most chatbots is their limited scope. Multi-agent-based systems offer approaches to solve problems in a cooperative manner following the “divide and conquer” paradigm. Consequently, it seems promising to design a multi-agent-based chatbot approach scaling beyond the scope of a single application context. To address this research gap, we propose a novel approach orchestrating well-established conversational assistants. We demonstrate and evaluate our approach using six chatbots, providing higher quality than competing artifacts.

The presence of computer vision technology is continually expanding into multiple application domains. An industry and an art form that is particularly attractive for the application of computer vision algorithms is ballet. Due to the well-codified poses, along with the challenges that exist within the ballet domain, automation for the ballet environment is a relevant research problem. The paper proposes a model called BaReCo, which allows for ballet poses to be recognised using computer vision methods. The model contains multiple computer vision pipelines which allows for the comparison of approaches that have not been widely explored in the ballet domain. The results have shown that the top-performing pipelines achieved an accuracy rate of 99.375% and an Equal Error Rate (EER) of 0.119% respectively. The study additionally produced a ballet pose dataset, which serves as a contribution to the ballet and computer vision community. By combining suitable computer vision methods, the study demonstrates that successful recognition of ballet poses can be accomplished.

Specific Language Impairment is a communication disorder regarding the mastery of language and conversation that impacts children. The system proposed aims to provide an alternative diagnosis method that does not rely on specific assessment tools. The system will accept a voice sample from the child and then detect indicators that differentiate individuals with specific language impairment from that voice sample. These indicators were based on the timbre and pitch characteristics of sound. Three different feature spaces are calculated, followed by derived features, with three different classifiers to determine the most accurate combination. The three feature spaces are Chroma, Mel-frequency cepstral coefficients (MFCC), and Tonnetz and the three classifiers are Support Vector Machines, Random Forest and a Recurrent Neural Network. MFCC, representing the timbre characteristic, was found to be the most accurate feature vector across all classifiers and Random Forest being the most accurate classifier across all feature spaces. The most accurate combination found was the MFCC feature vector with the Random Forest classifier with an accuracy level of 99%. The MFCC feature vector has the most features that are extracted giving the reason for the high accuracy. However, this accuracy decreases when the recorded word is three syllables or longer. The system proposed has proven to be a valid method that can detect SLI.

Digital transformation is radically changing the way companies conduct business and compete in established markets. In particular, a growing number of companies are switching from predominantly product-focused to platform-based business models. However, it remains unclear how these platforms should be designed to enable platform owners to maximize value capture. In this study, we investigated the interactions between platform openness and extension modularization and their influence on value capture in the context of digital transformation. To do so, we combined a case survey strategy with a configurational approach using fuzzy-set Qualitative Comparative Analysis. We found that there is no single condition necessary to achieve a high degree of value capture. Furthermore, our results show the importance of closedness and tight coupling of platforms and their applications. Finally, we confirmed the importance of interface conformance to high value capture. In addition, our results contribute to both theory and practice and provide implications for future research into the role of digital platforms in digital transformation.

Mergers and acquisitions (M&A) highly rely on the information technology systems integration (ISI) of two companies to realize the operation and resource synergy. Given the vast cost of M&A and the high failure rate of the post-merger ISI, various strategies, enforcement mechanisms, and potential determinants for the ISI success have been proposed. The capability of IS managers is one of the recognized factors that are critical to ISI success. This study extends the theory of relationship quality to explain how the relationship ability of IS personnel influences the outcomes of the ISI in the M&A context. The empirical survey of IS personnel shows that relationship ability improves not only relationship quality but also teamwork quality of the ISI. The findings shed light on the different functions of relationship capability in the managerial design and imply the significance of relational building ability as an important skill of IS personnel for achieving post-merger ISI success.

Indirect taxation is a significant source of livelihood for any nation. Tax evasion inhibits the economic growth of a nation. It creates a substantial loss of much needed public revenue. We design a method to single out taxpayers who evade indirect tax by dodging their tax returns. Towards this, we derive six correlation parameters (features), three ratio parameters from tax return statements submitted by taxpayers, and another parameter based on the business interactions among taxpayers using the TrustRank algorithm. Then we perform spectral clustering on taxpayers using these ten parameters (features). We identify taxpayers located at the boundary of each cluster by using kernel density estimation, which are further investigated to single out tax evaders. We applied our method on the iron and steel taxpayer’s data set provided by the Commercial Taxes Department, Government of Telangana, India.

An increasing number of organizations employ blockchain technology in their business process landscapes, especially when dealing with inter-organizational choreographies. Due to complex requirements with regard to data security and privacy in practice, however, no singular blockchain captures all use cases. Blockchains optimized for various levels of risk tolerance and confidentiality coexist in multi-chain environments, posing severe architectural challenges for blockchain-based Business Process Management Systems (BPMSs). Current state-of-the-art approaches lack the global perspective necessary, and focus on single-blockchain environments. In this paper, we alleviate these issues by developing a general architecture for multi-chain BPMSs for choreographies. We show the feasibility of our architecture by a prototypical implementation, and discuss future challenges using a concrete case study.

Business process management (BPM) supports organizations with their operational procedures. Traditional BPM focuses on structured processes but lacks support for flexible ones. Case management addresses this gap. The fragment-based case management (fCM) approach models processes as a set of repetitive, structured fragments. At run-time fragments are instantiated and composed to realize flexibility while data requirements synchronize their execution. So far, fCM does not consider data-to-data associations or object-to-fragment bindings. We investigate both by (i) extending fCM models and (ii) refining the execution semantics. For evaluation, we present a formal model based on colored Petri nets.

The success of online shopping combined with the convenience of home delivery leads to massive congestion in cities. CEP (Courier-Express-Parcels) service provider have increasing cost and service pressure, especially in the last mile parcel delivery. Therefore, we propose a process-based white-label last mile delivery platform as a smart city approach to counter congestion. It allows the consolidation of parcels on the last mile, considers customer preferences, and gives local carriers access to the parcel market. This platform is conceptualized based on insights from interviews and workshops with experts. Experiences from a pilot study are discussed.

Process choreography diagrams are the standard way of representing interactions between different parties to reach a common business goal. In order to enact choreographies in a trust-less environment, blockchain-based implementations have been proposed. They support trustful interactions, i.e., information generated on the blockchain during execution is trustworthy. However, existing solutions employ blockchain data that are bound to a single choreography. This paper proposes a novel approach to implement choreographies on the blockchain in a way that the generated data can be reused by different choreographies leading to cost reduction without sacrificing data integrity. The approach is evaluated in terms of feasibility and costs by developing a prototype based on the Ethereum blockchain.

A common challenge for improving business processes in large organizations is that business people in charge of the operations are lacking a fact-based understanding of the execution details, process variants, and exceptions taking place in business operations. While existing process mining methodologies can discover these details based on event logs, it is challenging to communicate the process mining findings to business people. In this paper, we present a novel methodology for discovering business areas that have a significant effect on the process execution details. Our method uses clustering to group similar cases based on process flow characteristics and then influence analysis for detecting those business areas that correlate most with the discovered clusters. Our analysis serves as a bridge between BPM people and business people, facilitating the knowledge sharing between these groups. We also present an example analysis based on publicly available real-life purchase order process data.

Using process mining actionable insights can be extracted from the event data stored in information systems. The analysis of event data may reveal many performance and compliance problems, and generate ideas for performance improvements. This is valuable, however, process mining techniques tend to be backward-looking and provide little support for forward-looking approaches since potential process interventions are not assessed. System dynamics complements process mining since it aims to capture the relationships between different factors at a higher abstraction level, and uses simulation to predict the effects of process improvement actions. In this paper, we propose a new approach to support the design of system dynamics models using event data. We extract a variety of performance parameters from the current state of the process using historical execution data and provide an interactive platform for modeling the performance metrics as system dynamics models. The generated models are able to answer “what-if” questions. Our experiments, using event logs including different relationships between parameters, show that our approach is able to generate valid models and uncover the underlying relations.

Designing and realizing data-driven business models (DDBMs) are key challenges for many enterprises and are recent research topics. While enterprise architecture (EA) modeling and management proved their potential value for supporting information technology-related projects, EA’s specific role in developing and realizing DDBMs is a new and rather unexplored research field. We conducted a systematic literature review on big data, business models, and EA to identify the potentials of EA support for developing and realizing DDBMs. We derived 42 EA concerns from the literature, structured along the dimensions of the business model canvas and the status of realization (as-is, to-be).

Understanding customer behaviors is deemed crucial to improve customers’ satisfaction and loyalty, which eventually is materialized in increased revenue. This paper tackles this challenge by using complex networks and multiple instance reasoning to examine the network structure of Customer Purchasing Behaviors. Our main contributions rely on a new multiple instance similarity to measure the interaction among customers based on the mutual information theory focuses on the customers’ bags, a new network construction approach involving customers, orders and products, and a new measure for evaluating its internal consistency. The simulations using 12 real-world problems support the effectiveness of our proposal.

Conversational agents (CAs) offer huge potential for service companies by creating social closeness and enabling fast and scalable communication with customers. However, investigation of utilitarian and especially hedonic value as driving motivations for using CAs is still nascent. We found social presence to be an important predictor for hedonic and utilitarian value and subsequent continuance intention. Moreover, we reveal customers’ continuance intention is determined primarily by hedonic value when expecting a CA, whereas focus shifts to utilitarian values if customers expect a human employee. With our results, CA services can be better tailored to customer needs and company service goals.

With the increasing number of similar web services nowadays, the need to satisfy the complex user requirements and locate relevant services remain necessary. As a complex and challenging task, many approaches have been proposed. Nevertheless, they totally neglect the contribution of the social dimension. The mix of two domains social computing with service oriented computing opens the door to new discovery schemes. It gives birth to a new notion Social Web Services. In fact, integrating the social aspect into web services can benefit them to become active entities that can collaborate, compete or substitute each other. In this paper, we present the second step of our social web service discovery model that operates on a bipartite graph with user-user, user-service and service-service relationships and employs new metrics to evaluate the ability of a user to help the service requester to satisfy his needs.

Nowadays, information describing navigation behaviour of internet users are used in several fields, e-commerce, economy, sociology and data science. Such information can be extracted from different knowledge bases, including business-oriented ones. In this paper, we propose a new model for the PageRank, CheiRank and 2DRank algorithm based on the use of clickstream and pageviews data in the google matrix construction. We used data from Wikipedia and analysed links between over 20 million articles from 11 language editions. We extracted over 1.4 billion source-destination pairs of articles from SQL dumps and more than 700 million pairs from XML dumps. Additionally, we unified the pairs based on the analysis of redirect pages and removed all duplicates. Moreover, we also created a bigger network of Wikipedia articles based on all considered language versions and obtained multilingual measures. Based on real data, we discussed the difference between standard PageRank, Cheirank, 2DRank and measures obtained based on our approach in separate languages and multilingual network of Wikipedia.

The advent of the Internet-of-Things (IoT) generates increasing data with the majority being gathered for a single purpose and staying unused after serving this purpose. With IoT platforms, cross-domain use cases, combining data from different sources, become possible. Accordingly, the need for marketplaces to trade data arises. This paper examines existing IoT platforms to frame the current opportunities for an IoT marketplace. In a second step, it analyzes the potentials of Distributed Ledger Technology (DLT) regarding transaction costs and efficiencies. In doing so, a classification regarding the functional distribution of IoT marketplaces is developed.

Fostering partnerships and generativity, the Industrial Internet of Things (IIoT) platforms change the way of value creation, enabling platform-based ecosystems. Platform Boundary Resources (BR) provide a recognized concept to foster third-party innovation, enabling various hardware- and software-developing companies to use the functionalities of the platform. Despite the high importance of BR to open the platform and foster the innovation, their design and quality aspects, as well as their influence on the satisfaction of complementors, remain under-researched. To understand how complementors value different BR in IIoT ecosystems, we conducted a complementor satisfaction survey, addressing developers in an IIoT ecosystem, who utilize various BR. The study is based on the case of the IIoT platform MindSphere, developed by Siemens. Our findings include the calculation of the weighted complementor satisfaction with BR. Adding the complementor satisfaction perspective to BR research, our study shows how to apply a structured quality improvement to the BR concept and supports platform providers, highlighting which BR should be in focus during the ecosystem development through the quality improvement of prioritized BR.

Cloud computing passed the hype cycle long ago and firmly established itself as a future technology since then. However, to utilize the cloud optimally, and therefore, as cost-efficiently as possible, a continuous monitoring is key to prevent an over- or under-commissioning of resources. However, selecting a suitable monitoring solution is a challenging task. Monitoring agents that collect monitoring data are spread across the monitored IT environment. Therefore, the possibility of vendor lock-ins leads to a lack of flexibility when the cloud environment or the business needs change. To handle these challenges, we introduce generic agent templates that are applicable to many monitoring systems and support a replacement of monitoring systems. Solution-specific technical details of monitoring agents are abstracted from and system administrators only need to model generic agents, which can be transformed into solution-specific monitoring agents. The transformation logic required for this process is provided by domain experts to not further burden system administrators. Furthermore, we introduce an agent lifecycle to support the system administrator with the management and deployment of generic agents.

The trend towards industry 4.0 amplifies existing data management challenges and requires suitable data governance and data quality measures. Although these topics have been previously discussed in literature, companies are still struggling to cope with the resulting challenges and fully exploit the benefits of industry 4.0. In this paper, we conducted a single case study in an automotive company. We exemplary used the material retrieval process in automotive manufacturing to uncover what challenges there are hindering the utilization of industry 4.0. We were able to identify six major challenges in the domains of data quality and data governance.

Additive Manufacturing is increasingly used in the industrial sector as a result of continuous development. In the Production Planning and Control (PPC) system, AM enables an agile response in the area of detailed and process planning, especially for a large number of plants. For this purpose, a concept for a PPC system for AM is presented, which takes into account the requirements for integration into the operational enterprise software system. The technical applicability will be demonstrated by individual implemented sections. The presented solution approach promises a more efficient utilization of the plants and a more elastic use.

Industry 4.0 use cases such as predictive maintenance and product quality control make it necessary to create, use and maintain a multitude of different machine learning models. In this setting, model management systems help to organize models. However, concepts for model management systems currently focus on data scientists, but do not support non-expert users such as domain experts and business analysts. Thus, it is difficult for them to reuse existing models for their use cases. In this paper, we address these challenges and present an architecture, a metadata schema and a corresponding model management platform.

Intelligent systems are technologically advanced machines that perceive and respond to the world around them. They can take many forms: facial recognition programs, personalized shopping suggestions, healthcare tools, etc. Research in intelligent systems faces numerous challenges, many of which relate to automatic reasoning. Intelligent systems’ knowledge bases are founded on facts and rules. Rules updates are essential to ensure that the system adapts to its environment evolution. In this paper, we aim to facilitate the automation of rule bases management by eliminating redundancies and handling circularity. This research work is part of the proposition of an approach for automating the management of rule bases. Our method is based on dependency relationships that may exist between the rules. The experimentation results show that our proposition succeeded in eliminating redundancies and detecting a great number of cycles.

Automatic Emotion recognition is renowned for being a difficult task, even for human intelligence. Due to the importance of having enough data in classification problems, we introduce a framework developed with the purpose of generating labeled audio to create our own database. In this paper we present a new model for audio-video emotion recognition using Transfer Learning (TL). The idea is to combine a pre-trained high level feature extractor Convolutional Neural Network (CNN) and a Bidirectional Recurrent Neural Network (BRNN) model to address the issue of variable sequence length inputs. Throughout the design process we discuss the main problems related to the high complexity of the task due to its inherent subjective nature and, on the other hand, the important results obtained by testing the model on different databases, outperforming the state-of-the-art algorithms in the SAVEE [3] database. Furthermore, we use the mentioned application to perform precision classification (per user) into low resources real scenarios with promising results.