DYFRAM: dynamic fragmentation and replica management in distributed database systems

In distributed database systems, tables are frequently fragmented and replicated over a number of sites in order to reduce network communication costs. How to fragment, when to replicate and how to allocate the fragments to the sites are challenging problems that has previously been solved either by static fragmentation, replication and allocation, or based on a priori query analysis. Many emerging applications of distributed database systems generate very dynamic workloads with frequent changes in access patterns from different sites. In such contexts, continuous refragmentation and reallocation can significantly improve performance. In this paper we present DYFRAM, a decentralized approach for dynamic table fragmentation and allocation in distributed database systems based on observation of the access patterns of sites to tables. The approach performs fragmentation, replication, and reallocation based on recent access history, aiming at maximizing the number of local accesses compared to accesses from remote sites. We show through simulations and experiments on the DASCOSA distributed database system that the approach significantly reduces communication costs for typical access patterns, thus demonstrating the feasibility of our approach.     

Domain science and engineering from computer science to the sciences of informatics. Part I: Engineering

In this paper, we wish to advocate that departments of computer science put emphasis on teaching programming and software engineering based on formal methods; and more emphasis on research into formal methods for the trustworthy development of software. We also wish to advocate that the concepts of domain science and domain engineering become an indispensable part of the science of informatics and of software engineering.     

Eigenvectors of directed graphs and importance scores: dominance,T-Rank,and sink remedies

We study the properties of the principal eigenvector for the adjacency matrix (and related matrices) for a general directed graph. In particular—motivated by the use of the eigenvector for estimating the “importance” of the nodes in the graph—we focus on the distribution of positive weight in this eigenvector, and give a coherent picture which builds upon and unites earlier results. We also propose a simple method—“T-Rank”—for generating importance scores. T-Rank generates authority scores via a one-level, non-normalized matrix, and is thus distinct from known methods such as PageRank (normalized), HITS (two-level), and SALSA (two-level and normalized). We show, using our understanding of the principal eigenvector, that T-Rank has a much less severe “sink problem” than does PageRank. Also, we offer numerical results which quantify the “tightly-knit community” or TKC effect. We find that T-Rank has a stronger TKC effect than PageRank, and we offer a novel interpolation method which allows for continuous tuning of the strength of this TKC effect. Finally, we propose two new “sink remedies”, i.e., methods for ensuring that the principal eigenvector is positive everywhere. One of our sink remedies (source pumping) is unique among sink remedies, in that it gives a positive eigenvector without rendering the graph strongly connected. We offer a preliminary evaluation of the effects and possible applications of these new sink remedies.     

Optimal Sparse Matrix Dense Vector Multiplication in the I/O-Model

We study the problem of sparse-matrix dense-vector multiplication (SpMV) in external memory. The task of SpMV is to compute y:=Ax, where A is a sparse N×N matrix and x is a vector. We express sparsity by a parameter k, and for each choice of k consider the class of matrices where the number of nonzero entries is kN, i.e., where the average number of nonzero entries per column is k.     

A visual digital library approach for time-oriented scientific primary data

Digital Library support for textual and certain types of non-textual documents has significantly advanced over the last years. While Digital Library support implies many aspects along the whole library workflow model, interactive and visual retrieval allowing effective query formulation and result presentation are important functions. Recently, new kinds of non-textual documents which merit Digital Library support, but yet cannot be fully accommodated by existing Digital Library technology, have come into focus. Scientific data, as produced for example, by scientific experimentation, simulation or observation, is such a document type. In this article we report on a concept and first implementation of Digital Library functionality for supporting visual retrieval and exploration in a specific important class of scientific primary data, namely, time-oriented research data. The approach is developed in an interdisciplinary effort by experts from the library, natural sciences, and visual analytics communities. In addition to presenting the concept and to discussing relevant challenges, we present results from a first implementation of our approach as applied on a real-world scientific primary data set. We also report from initial user feedback obtained during discussions with domain experts from the earth observation sciences, indicating the usefulness of our approach.     

Control strategies used by experienced marine navigators: observation of verbal conversations during navigation training

This article investigates the relationship between the control situation framework, as presented by Bjørkli et al. (Cogn Technol Work 9:67–80, 2007), and observed operator action and communication in high-speed craft operation in the Norwegian Royal Navy. The reported data include two series of events during sailing in the Norwegian archipelago where uncertainty of ship position, challenges of manoeuvring and navigation strategies are addressed. The results indicate that the navigators perform actions and communicate in accordance with the key features of the control situation framework, and thus adapt to the interrelation between control possibilities and control requirements. The navigators were guided by explicit knowledge of the functional characteristics of psychical and temporal dynamics, which were actualized in manoeuvring. It indicates that the control situation framework can be used to describe the models the operators use.     

Predicting Articulated Human Motion from Spatial Processes

We present a probabilistic interpretation of inverse kinematics and extend it to sequential data. The resulting model is used to estimate articulated human motion in visual data. The approach allows us to express the prior temporal models in spatial limb coordinates, which is in contrast to most recent work where prior models are derived in terms of joint angles. This approach has several advantages. First of all, it allows us to construct motion models in low dimensional spaces, which makes motion estimation more robust. Secondly, as many types of motion are easily expressed in spatial coordinates, the approach allows us to construct high quality application specific motion models with little effort. Thirdly, the state space is a real vector space, which allows us to use off-the-shelf stochastic processes as motion models, which is rarely possible when working with joint angles. Fourthly, we avoid the problem of accumulated variance, where noise in one joint affects all joints further down the kinematic chains. All this combined allows us to more easily construct high quality motion models. In the evaluation, we show that an activity independent version of our model is superior to the corresponding state-of-the-art model. We also give examples of activity dependent models that would be hard to phrase directly in terms of joint angles.     

Incremental Linear Discriminant Analysis Using Sufficient Spanning Sets and Its Applications

This paper presents an incremental learning solution for Linear Discriminant Analysis (LDA) and its applications to object recognition problems. We apply the sufficient spanning set approximation in three steps i.e. update for the total scatter matrix, between-class scatter matrix and the projected data matrix, which leads an online solution which closely agrees with the batch solution in accuracy while significantly reducing the computational complexity. The algorithm yields an efficient solution to incremental LDA even when the number of classes as well as the set size is large. The incremental LDA method has been also shown useful for semi-supervised online learning. Label propagation is done by integrating the incremental LDA into an EM framework. The method has been demonstrated in the task of merging large datasets which were collected during MPEG standardization for face image retrieval, face authentication using the BANCA dataset, and object categorisation using the Caltech101 dataset.     

Defining syntax-directed translations by tree bimorphisms

We introduce a class of tree bimorphisms that define exactly the translations performed by syntax-directed translation schemata. We also show that these “quasi-alphabetic” tree bimorphisms preserve recognizability, and that their class is closed under composition and inverses.