Ever-increasing amounts of data and requirements to process them in real time lead to more and more analytics platforms and software systems being designed according to the concept of stream processing. A common area of application is the processing of continuous data streams from sensors, for example, IoT devices or performance monitoring tools. In addition to analyzing pure sensor data, analyses of data for groups of sensors often need to be performed as well. Therefore, data streams of the individual sensors have to be continuously aggregated to a data stream for a group. Motivated by a real-world application scenario, we propose that such a stream aggregation approach has to allow for aggregating sensors in hierarchical groups, support multiple such hierarchies in parallel, provide reconfiguration at runtime, and preserve the scalability and reliability qualities induced by applying stream processing techniques. We propose a stream processing architecture fulfilling these requirements, which can be integrated into existing big data architectures. We present a pilot implementation of such an extended architecture and show how it is used in industry. Furthermore, in experimental evaluations we show that our solution scales linearly with the amount of sensors and provides adequate reliability in the case of faults.

Supervised learning classification models use labeled data to train models on a discrete form for generating predictions. A major challenge addressed in this paper is training a machine learning model to the recognition of a pattern data perspective of the original datasets and privacy-preserving datasets to improve predictive models. The model training process, the training datasets, and validation datasets are mixed with data and privacy-preserving data cause overfitting from high variance in the machine learning algorithm. This paper addresses a k-Nearest Neighbor algorithm to build models, apply an automated hyperparameter tuning method to determine the optimal parameters based on the characteristics before the training process of a large volume datasets. Evaluating the model to achieve goals based on a high score of accuracy results on quality prediction and performance models. The experiments from our real datasets and the UCI machine learning repository show the best method for all of the training data and conduct difference experiments for improving accuracy, feasibility, correctness and reliability of the scheme.

A main problem for electronic commerce, particularly for business-to-business applications, lies in the need for the involved information systems to meaningfully exchange information. Domain-specific standards may be used to define the semantics of common terms. However, in practice it is not easy to find those domain-specific standards that are detailed and stable enough to allow for real interoperability. Therefore, we propose an architecture that allows for incremental construction of a shared repository including a multilingual thesaurus, which is used in a business communication language. Communicating information systems then refer to the common thesaurus while exchanging messages. Our emphasis is be on separating semantics (in the thesaurus) and syntax (in XML). Therefore, our extensibility is not only that of XML, but also the extensibility of the semantics that is modeled in the shared repository. The business communication language XLBC is presented and how it can be used in electronic commerce applications. XLBC message patterns and conversation protocols are stored in the shared repository as well.