Contributions in Merlin

null

null

Since its introduction in 1971, the B-tree has become the dominant index structure in database systems. Conventional wisdom dictated that the use of a B-tree index or one of its descendants would typically lead to good results. The advent of XML-data, column stores, and the recent resurgence of typed-graph (or triple) stores motivated by the Semantic Web has changed the nature of the data typically stored. In this paper we show that in the case of triple-stores the usage of B-trees is actually highly detrimental to query performance. Specifically, we compare on-disk query performance of our triple-based Hexastore when using two different B-tree implementations, and our simple and novel vector storage that leverages offsets. Our experimental evaluation with a large benchmark data set confirms that the vector storage outperforms the other approaches by at least a factor of four in load-time, by approximately a factor of three (and up to a factor of eight for some queries) in query-time, as well as by a factor of two in required storage. The only drawback of the vector-based approach is its time-consuming need for reorganization of parts of the data during inserts of new triples: a seldom occurrence in many Semantic Web environments. As such this paper tries to reopen the discussion about the trade-offs when using different types of indices in the light of non-relational data and contribute to the endeavor of building scalable and fast typed-graph databases.

What is the nature of imperfections in the market for liquidity? Studying bidder level data from European Central Bank (ECB) repo auctions, we find that this market appears to be informationally efficient in the sense that participants do not have private information about future short-term rates. However, auction allocations affect banks' subsequent behavior in a way that is consistent with a degree of allocational and operational inefficiency. Also, large bidders appear to have better access to the interbank market than small ones. Finally, the evidence suggests that the ECB uses collateral haircuts that do not equilibrate opportunity costs.

This article develops a contest model to compare social welfare in homogeneous leagues in which all clubs maximize identical objective functions with mixed leagues in which clubs maximize different objective functions. We show that homogeneous leagues in which all clubs are profit maximizers dominate all other leagues. Mixed leagues in which small-market clubs are profit maximizers and large-market clubs are win maximizers (type-I mixed leagues) are dominated by all other leagues. From a welfare perspective, large-market clubs win too often in (purely) win-maximizing and type-I mixed leagues; whereas, small-market clubs win too many games in (purely) profit-maximizing leagues and in mixed leagues in which largemarket clubs are profit maximizers and small-market clubs are win maximizers (type-II mixed leagues). These results have important policy implications: Social welfare will increase if clubs are reorganized from non-profit member associations to profit-maximizing corporations. Moreover, we show that revenue sharing decreases (increases) social welfare in mixed (homogeneous) leagues.

This study investigates the prevalence and extent of altruism by examining the relationship between parents’ and their adult children’s subjective well-being in a data set extracted from the German Socio-Economic Panel. To segregate the share of parents with altruistic preferences from those who are selfish, we estimate a finite mixture regression model. We control for various sources of potential bias by taking advantage of the data’s panel structure. To validate our modeling approach, we show that predicted altruists indeed make higher average transfer payments.

Television via peer-to-peer (P2P) networks (P2PTV) and television on mobile devices (Mobile TV) are promising technologies for the future. Attempts to combine the two techniques into Mobile P2PTV are discussed in scientific papers, but no real implementations have been successfully completed. This work designs and implements a hybrid Mobile P2PTV prototype for devices running on Google Android. The prototype connects to the P2PTV application LiveShift, developed at the University of Zurich. The prototype does not participate in the P2P network himself, but connects to a participating peer and retrieves the video data from there. It is able to perform the same functions as the peer is able to perform by remote controlling the peer. The goal is to develop a client for a Google Android phone that can connect to a running LiveShift application and control it for the reception and displaying of video channels. The client should also be able to fetch an electronic program guide. A transmission protocol should also be designed that is used to remote control the LiveShift instance. The design proposes to split the problem in a LiveShift, a client and a transmission protocol part. While the LiveShift part provides a general communication interface and the logic to receive and process messages from the client, the transmission protocol is responsible for transmitting the messages and defining how LiveShift has to react on receipt. The client provides the function to watch all LiveShift channels while it is connected to a running LiveShift instance. The client can remote control LiveShift to perform the same actions as LiveShift can. The implementation is done in Java. For the client application the Android Java framework is used. Not all aimed goals have been achieved. The EPG function is working, but cannot display program details. The transmission protocol is implemented successfully and controlling LiveShift remotely works, but the display of video channels from the LiveShift network is not possible. The Android device cannot decode and display the received video data successfully. This unsolved problem can be investigated in future work on this topic.

null