Industry 4.0 has begun to shape the way organizations operate by emphasizing the need for a duality between physical machines and sensors, and the (big) data they generate, exchange and use. Manufacturing is one of several industries which is expected to be impacted by this technological revolution. Increasing the information flows and integration of systems within organizations, and along the supply chain is considered one of the main challenges that needs to be addressed by these organizations. One approach for addressing this challenge is to investigate how this abundance of (big) operational data can be used in combination with IT-driven design approaches, such as Enterprise Architecture. Therefore, in this paper we propose our vision for Enterprise Architecture 4.0, i.e. an extended Enterprise Architecture approach for the context of Industry 4.0, and we give an account of our (work-in-progress) efforts to design a model management and analytics software platform supporting this vision. The usage of the software tool is exemplified with the help of a case study with an organization that develops IT and automation systems for the husbandry sector.

To facilitate the sharing and re-use of data in scientific studies we propose an automated technique for annotating operation results. The annotated output has to preserve, as much as possible, the properties of the input annotations. The preservation of properties is achieved by taking into account operation properties. Property preservation is evaluated with information theory metrics.

The increasing data volume and highly complex models used in different domains make it difficult to debug models in cases of anomalies. Data provenance provides scientists sufficient information to investigate their models. In this paper, we propose a tool which can infer fine-grained data provenance based on a given script. The tool is demonstrated using a hydrological model. The tool is also tested successfully handling other scripts in different contexts.

Structured prediction has wide applications in many areas. Powerful and popular models for structured prediction have been developed. Despite the successes, they suffer from some known problems: (i) Hidden Markov models are generative models which suffer from the mismatch problem. Also it is difficult to incorporate overlapping, non-independent features into a hidden Markov model explicitly. (ii) Conditional Markov models suffer from the label bias problem. (iii) Conditional Random Fields (CRFs) overcome the label bias problem by global normalization. But the global normalization of CRFs can be expensive which prevents CRFs from applying to big data. In this paper, we propose the Empirical Co-occurrence Rate Networks (ECRNs) for sequence labeling. ECRNs are discriminative models, so ECRNs overcome the problems of HMMs. ECRNs are also immune to the label bias problem even though they are locally normalized. To make the estimation of ECRNs as fast as possible, we simply use the empirical distributions as the estimation of parameters. Experiments on two real-world NLP tasks show that ECRNs reduce the training time radically while obtain competitive accuracy to the state-of-the-art models.

Data provenance allows scientists to validate their model as well as to investigate the origin of an unexpected value. Furthermore, it can be used as a replication recipe for output data products. However, capturing provenance requires enormous effort by scientists in terms of time and training. First, they need to design the workflow of the scientific model, i.e., workflow provenance, which requires both time and training. However, in practice, scientists may not document any workflow provenance before the model execution due to the lack of time and training. Second, they need to capture provenance while the model is running, i.e., fine-grained data provenance. Explicit documentation of fine-grained provenance is not feasible because of the massive storage consumption by provenance data in the applications, including those from the geoscience domain where data are continuously arriving and are processed. In this paper, we propose an inference-based framework, which provides both workflow and fine-grained data provenance at a minimal cost in terms of time, training, and disk consumption. Our proposed framework is applicable to any given scientific model, and is capable of handling different model dynamics, such as variation in the processing time as well as input data products arrival pattern. Our evaluation of the framework in a real use case with geospatial data shows that the proposed framework is relevant and suitable for scientists in geoscientific domain.

Scientists require provenance information either to validate their model or to investigate the origin of an unexpected value. However, they do not maintain any provenance information and even designing the processing workflow is rare in practice. Therefore, in this paper, we propose a solution that can build the workflow provenance graph by interpreting the scripts used for actual processing. Further, scientists can request fine-grained provenance information facilitating the inferred workflow provenance. We also provide a guideline to customize the workflow provenance graph based on user preferences. Our evaluation shows that the proposed approach is relevant and suitable for scientists to manage provenance.

Two orchestrated processes interacting with each other have to maintain their own states. Messages are used to synchronize states between orchestrated processes. Server crash and network failure may result in loss of messages and therefore result in a state change performed by only one party. Thus, the states of the parties are no longer synchronized, resulting in state inconsistencies and in worst case deadlocks. In this paper, we propose a mechanism for guaranteed state synchronization of orchestrated processes with system and network failures. Our mechanism is based on interaction patterns and process transformations. The basic idea is to redesign the original processes into their state synchronization-enabled counterparts via process transformations that can be automated. The transformation mechanism is formalized based on Colored Petri Nets. We present the formal proof of the correctness of our mechanism and give the overhead analysis to illustrate its practicability.

Scientific data are often annotated based on their properties, which are not maintained during further data processing. Not maintaining annotations results in loss of information. Decisions made on such incomplete information may be wrong. In this paper the problem of propagating annotations along a data processing chain is formulated. In particular, an annotation of a data element is an identification that this data element exhibits a specific property. The propagation of this property from the input of an operation to its output is called the identification problem. In this paper the identification problem is described as a clustering problem.

In stream data processing, data arrives continuously and is processed by decision making, process control and e-science applications. To control and monitor these applications, reproducibility of result is a vital requirement. However, it requires massive amount of storage space to store fine-grained provenance data especially for those transformations with overlapping sliding windows. In this paper, we propose techniques which can significantly reduce storage costs and can achieve high accuracy. Our evaluation shows that adaptive inference technique can achieve almost 100% accurate provenance information for a given dataset at lower storage costs than the other techniques. Moreover, we present a guideline about the usage of different provenance collection techniques described in this paper based on the transformation operation and stream characteristics.

During the last years a new generation of adaptive Process-Aware Information Systems (PAIS) has emerged, which enables dynamic process changes at runtime, while preserving PAIS robustness and consistency. Such adaptive PAIS allow authorized users to add new process activities, to delete existing activities, or to change pre-defined activity sequences during runtime. Both this runtime flexibility and process configurations at build-time, lead to a large number of process variants being derived from the same process model, but slightly differing in structure due to the applied changes. Generally, process variants are expensive to configure and difficult to maintain. This paper presents selected results from our MinAdept project. In particular, we provide a clustering algorithm that fosters learning from past process changes by mining a collection of process variants. As mining result we obtain a process model for which average distance to the process variant models becomes minimal. By adopting this process model as reference model in the PAIS, need for future process configuration and adaptation decreases. We have validated our clustering algorithm by means of a case study as well as comprehensive simulations. Altogether, our vision is to enable full process lifecycle support in adaptive PAIS.

Inter-organizational cooperations are complex in terms of coordination, agreements, and value creation for involved partners. When managing complex cooperations, it is vital to maintain models describing them. Changing one model to regain consistency with the running system might result in new inconsistencies. As a consequence, this maintenance phase grows in complexity with increasing number of models. In this context, challenges are to ensure consistency at design time and to monitor the system at runtime, i.e., at design time, consistency between different models describing the cooperation needs to be ensured. At runtime, behavior of the software system needs to be compared with its underlying models. In this paper, we propose a structured and model-independent method that supports ensuring and maintaining consistency between running system and underlying models for inter-organizational cooperations.

Recently, a new generation of adaptive process management technology has emerged, which enables dynamic changes of composite services and process models respectively. This, in turn, results in a large number of process variants derived from the same process model, but differing in structure due to the applied changes. Since such process variants are expensive to maintain, the process model should be evolved accordingly. In this context, we need to know which activities have been more often involved in process adaptations than others, such that we can focus on them when reconfiguring the process model. This paper provides two approaches for ranking activities according to their involvement in process adaptations. The first one allows to precisely rank the activities, but is expensive to perform since the algorithm is at NP level. We therefore provide as alternative an approximation ranking algorithm which computes in polynomial time. The performance of the approximation algorithm is evaluated and compared through a simulation of 3600 process models. Statistical significance tests indicate that the performance of the approximation ranking algorithm does not depend on the size of process models, i.e., our algorithm can scale up.

Recently, a new generation of adaptive Process-Aware Information Systems (PAIS) has emerged, which allows for dynamic process and service changes (e.g., to insert, delete, and move activities and service executions in a running process). This, in turn, has led to a large number of process variants derived from the same model, but differing in structure due to the applied changes. Generally, such process variants are expensive to configure and difficult to maintain. This paper provides a sophisticated approach which fosters learning from past process changes and allows for mining process variants. As a result we obtain a generic process model for which the average distance between this model and the respective process variants becomes minimal. By adopting this generic model in the PAIS, need for future process configuration and adaptation decreases. We have validated the proposed mining method and implemented it in a powerful proof-of-concept prototype.

An essential issue in peer-to-peer data management is to keep data highly available all the time. A common idea is to replicate data hoping that at least one replica is available when needed. However, due to churns, the number of created replicas could be not sufficient for guaranteeing the intended data availability. If the number of replicas is computed according to the lowest expected peer availability (a classical case), but the expectation were too high, then the peer availability after a churn could be too low, and the system could not be able to recover the requested data availability. The paper is a continuation of previous work (Knezevic et al., 2006) and presents a replication protocol that delivers a configured data availability guarantee, and is resistant to, or recovers fast from churns. The protocol is based on a distributed hash table (DHT), measurement of peer online probability in the system, and adjustment of the number of replicas accordingly. The evaluation shows that we are able to maintain or recover the requested data availability during or shortly after stronger or weaker churns, and at the same time the storage overhead is close to the theoretical minimum

Process-aware information systems have to be frequently adapted due to business process changes. One important challenge not adequately addressed so far concerns the evolution of process choreographies. If respective modifications are applied in an uncontrolled manner, inconsistencies or errors might occur in the sequel. In particular, modifications of private processes performed by a single party may affect the implementation of the private processes of partners as well. In this paper we sketch a framework that allows process engineers to detect how changes of private processes may affect related public views and - if so - how they can be propagated to the public and private processes of partners. Our approach exploits the semantics of the applied changes in order to automatically determine the adaptations necessary for the partner processes.

The existing service discovery infrastructure with UDDI as the de facto standard, is limited in that it does not support more complex searching based on matching business processes. Two business processes match if they agree on their simple services, their processing order as well as any mandatory or optional requirements for the service. This matching semantics can be formalized by modelling business processes as annotated finite state automata (aFSAs) and deciding emptiness of the intersection aFSA. Computing the intersection of aFSAs and deciding emptiness are computationally expensive, being more than quadratic on the number of states and transitions, thus does not scale for large service repositories. This paper presents an approach for indexing and matching business processes modeled as aFSAs, for the purpose of service discovery. Evaluation of this approach shows a performance gain of several orders of magnitude over sequential matching and a linear complexity with regard to the data set size

P2P data stores achieve high data availability by introducing replication of data. As a consequence, searching data may require to follow several routing paths introducing additional communication. A formal model of this additional communication is introduced and evaluated based on a hypercube-based P2P data store

Complex services are composed of simple services which typically need to be processed in a particular order. Two complex services only match if they agree on both, their simple services and their processing order. This matching semantics can be formalized by means of modelling complex services as finite state automata (FSAs), and analysing the intersection of the FSAs. However, computing the intersection of FSAs is computationally expensive, and thus, does not scale for large service repositories. This paper presents an approach for indexing and matching complex services using an abstraction that transforms the underlying FSA via its grammar into a form that can be indexed using available index mechanisms. Evaluation of this approach shows a performance gain of several orders of magnitude as compared to sequential matching.

Service composition based on state dependent services is a challenge if it is done in a decentralized way that is without a centralized coordinating partner knowing all involved parties. In particular, the challenge is the combination of services to a composite service, such that every party involved considers its view of the composite service to be acceptable. In the paper a protocol is proposed which incrementally derives proposals for composite services. These proposals are then evaluated by the involved parties locally and a consensus on the local decisions on accepting or rejecting the proposed composite service is derived.

Web services advocate loosely coupled systems, although current loosely coupled applications are limited to stateless services. The reason for this limitation is the lack of a method supporting matchmaking of state dependent services exemplarily specified in BPEL. In particular, the sender's requirement that the receiver must support all possible messages sent at a certain state are not captured by models currently used for service discovery. Annotated deterministic finite state automata provide this expressiveness. In this paper the transformation of a local process specification given in BPEL to annotated deterministic finite state automata is presented.