Die GI-Regionalgruppe Münsterland lädt zu einem Sondertermin ein zum Vortrag von Dr. C. Mohan, IBM Almaden Research Center, San Jose, USA

TITEL: New Era in Distributed Computing with Blockchains and Databases

ABSTRACT:

In this talk we will present a framework for processing knowledge inconsistency in a collective using consensus computing. The model of knowledge inconsistency has been built with several inconsistency measures. We will show that inconsistency degree of a collective has an essential impact on the quality of collective knowledge. In addition to inconsistency degree, the number of collective members, the problem of collective knowledge increase, additional knowledge elements, multi-stage consensus choice are also investigated. In this short work we present briefly our latest results on this subject.

The increasing amount of data being generated on a daily basis has empowered the development of novel products and services. However, this data deluge has clear impacts on our everyday lives, turning usual activities, such as selecting destinations for vacation, choosing a movie on a streaming platform, or purchasing a suitable car, into puzzling situations in which there is no standard method to find an appropriate decision. To solve problems like these, personalized applications (from Web search engines to Recommender Systems) allow people to focus on useful options considering individual or group interests. Nonetheless, even when supported by systems like these, one can often notice that interesting options were not covered in a system's response. This is frequently caused by the fact that, when dealing with such decision problems, one frequently only has a vague idea about which characteristics are expected to be found in the final solution.  In this talk, the concept of Semiotic Machine will be introduced, a query refinement process for tackling uncertainty and vagueness in database queries. Moreover, a small case will be presented for materializing the ideas and for guiding further research avenues.

Humanitarian aid organizations respond to disasters by delivering relief items in highly complex and dynamic environments. To plan and execute logistics operations therein, it is crucial to build and maintain situation awareness. That is, to perceive the relevant elements of the environment, to comprehend their meaning and to project their status in the near future. Traditional data sources, such as direct observation by assessment teams, are limited in terms of speed and coverage. Crowdsourcing has emerged as a promising paradigm, process and technology in crisis management to complement traditional data sources and improve situation awareness. Until now, however, humanitarian decision-making, situation assessment and crowdsourcing initiatives have been largely disconnected. Before crowdsourcing can be effectively utilized to support humanitarian logistics decision-making, it is necessary (1) to be clear and what kinds of logistics-related data have to be collected, (2) to embed the crowdsourcing process in professional assessment practices, and (3) to provide adequate information systems support. The speaker will present research results for each of these challenges, including a category system for logistics-related infrastructure and resources, a crowdsourcing model and method, and the instantiating prototype GDACSmobile.

Dealing with data necessitates understanding data, which can be facilitated by analyzing its metadata. This metadata is often not readily available, and thus, needs to be extracted and collected. This activity is called data profiling and is well established and researched in the literature. However, what is missing is a structured description of a holistic data profiling process that brings the various tools, techniques and algorithms together. The goal is to achieve a guideline that helps a user get faster and more precise insights into his data.

The use of cryptocurrencies, like Bitcoin, is steadily growing. Bitcoins are used in e-commerce to purchase both legal and illegal goods, they are transferred and traded, and investors have brought a significant amount of capital to this virtual currency. While the technical aspects of the system are meanwhile well established, many legal questions remain open. Particularly, the de-central organization of the peer-to-peer network makes it difficult to apply existing law. The talk gives an overview of selected legal issues concerning Bitcoins under civil, criminal, and regulatory law. It points out possible solutions approaches and remaining research gaps.