Contributions in Merlin

null

This paper studies the effect of a large-scale policy change in the Austrian disability insurance program, which tightened eligibility criteria for men above a certain age. Using administrative data on the universe of Austrian private-sector employees, the results of difference-in-difference regressions suggest a substantial and statistically significant decline in disability enrollment of 6 to 7.4 percentage points and an increase in employment of 1.6 to 3.4 percentage points. The policy change had important spillover effects into the unemployment and sickness insurance program. Specifically, the share of individuals receiving unemployment benefits increased by 3.5 to 3.9 percentage points, and the share receiving sickness insurance benefits, by roughly 0.7 percentage points.

New demands for analyzing and working with large data sets establish new challenges for computation models, especially when dealing with Semantic Web information. Signal/Collect proposes an elegant model for applying graph algorithms on various data sets. However, a distributed feature for horizontally scaling and processing large volumes of data is missing. This thesis analyzes existing graph computation models and compares distributed message- passing frameworks for proposing an integrated Distributed Signal/Collect solution that tries to solve the problem of limited scalability. We successfully show that it is possible to implement distributed mechanisms using the Actor Model, although with some caveats. We also propose future works in an attempt to further enhance our solution.

In the scope of the SciMantic project a Knowledge Sharing System (KSS) is defined, that enables its users to share knowledge among each other. The knowledge within the system is divided in Knowledge Units. There is a prototype of such a KSS. It is implemented on top of Apache Clerezza, an OSGi based platform written in Java. KSS uses a structured P2P network to share the Knowledge Units among the users, where a KSS instance acts as a node or peer in this network. But until now the prototype does not have a notification mechanism that would allow the users to be notified if a Knowledge Unit of their interest is updated or created. This thesis presents three possible mechanisms on how to expand the KSS to enable users to be notified. One of these possibilities was chosen, for reasons explained in the paper, to be implemented and the implementation was evaluated in the scope of performance and scalability. The communication between the notifying node and the subscriber occurs through HTTP methods. And the tests show that the implemented notification mechanism is scalable but if a node has to handle too many requests at the same time, a java.net.SocketException is thrown. A reason for this could be that the machine has too many open files, which causes such an exception.

In every application that uses a database server, short response times for queries are desired. Semantic caching is a well-known approach to this issue and is based on caching executed queries locally for a future usage. An architecture of an advanced semantic caching system has been designed at the University of Zurich in order to further improve the performance. This architecture additionally uses other clients’ caches and distributes an index of all local caches via a P2P network. The cooperative semantic caching approach has been implemented in a C++ prototype application called CoopSC. Since the use of CoopSC is also desirable in Java programs, the task of this assignment is to develop a JDBC driver for CoopSC. The key question of this assignment is how to bridge the C++ and Java worlds in order to create a JDBC driver that makes the CoopSC functionality available in Java. In the course of this assignment JNI has been chosen to fulfill this task and a JDBC driver could be implemented. An evaluation of this approach has been done by developing an adequate benchmark application and running tests with it. In these tests the response times for answering queries by CoopSC directly was compared with answering them via the JDBC driver. The evaluation has shown that the JDBC driver is creating an overhead which is mainly caused by the use of JNI. Furthermore, a GUI has been developed in this assignment which is running on top of the JDBC driver.

The principle introduced in this thesis enables a robot to learn reflex-like behaviours automatically. This is done by performing twitches with a tendon-driven pendulum robot equipped with several sensor modalities (force, length- and tactile sensors) and leads to the emergence of four behaviours (equivalent to myotatic reflex, reverse myotatic reflex, reciprocal inhibition reflex and withdrawal reflex). The reflexes emerge from the structure of the robot with the help of a correlation-based learning scheme. The thesis describes the four reflexes, the robot including its sensors, the learning procedure and the production of the reflexes.

Numerous software analysis techniques and tools have been proposed throughout the years. They consume information like version control, bug and issue tracking, to assess software quality. However, because of their platform dependency or special format of input and output data they lack synergies. Hence, their results are hard to compare or relate. The goal of the SOFAS project is to overcome these issues by introducing a distributed and collaborative software analysis platform. The platform offers analyses in form of web services, so that they can be easily accessed from any location in a remote way. This thesis contributes to the SOFAS project, enriching its functionality. A set of new analysis services were developed and a prototype of a simple workflow engine for the composition and the automated analysis execution. These tools were tested in a case study of a thorough analysis of a real project. Finally, the limitations of them and the SOFAS platform as a whole, together with some thoughts on possible future work, will be addressed in this report.

A main drawback in today's upper limb prostheses is the lack of sensory feedback, requiring patients to operate the prosthesis under sight control. A stimulation device, using transcutaneous electrical nerve stimulation has been previously developed at the University of Zurich. The goal was to embed this stimulator in a myoelectric prosthetic control environment for further investigation on the benefit this kind of feedback is able to provide for prosthetic operation. A modular framework was developed, integrating EMG sensors, a robotic hand and the stimulator. The prosthetic hand's force and position information was output onto the user's lower back through one channel each. Experiments showed that the stimulator was able to return perceivable information about the hand's status.

Am Beispiel des Untergangs der Lusitania und der Titanic wird analysiert, ob und in welchem Maße finanzielle und physische Macht sowie soziale Normen über Leben und Tod entscheiden. In einem quasi-natürlichen Experiment werden multivariate Probit-Schätzungen von öffentlich verfügbaren Sekundärdaten der Schiffsuntergänge durchgeführt. Die Analyse kommt zum Ergebnis, dass es im Wesentlichen von der Zeitspanne zwischen der Beschädigung des Schiffes und seinem Untergang abhängt, welche Rolle physische Stärke, gesellschaftlicher Status oder soziale Normen in lebensbedrohenden Situationen spielen. In zeitlich eng begrenzten Extremsituationen verdrängen Angst und Stress wertbezogenes, rationales Handeln. Es kommt zu einem rücksichtslosen Kampf ums eigene Überleben. Bleibt jedoch in Empfinden und Wahrnehmung der Betroffenen ein größerer Zeitraum bis zum endgültigen Versinken des Schiffes, bestimmen in stärkerem Maße soziale und ethische Werte das Verhalten der Menschen.

null