Finding the Most Descriptive Substructures in Graphs with Discrete and Numeric Labels

Many graph datasets are labelled with discrete and numeric attributes. Most frequent substructure discovery algorithms ignore numeric attributes; in this paper we show how they can be used to improve search performance and discrimination. Our thesis is that the most descriptive substructures are those which are normative both in terms of their structure and in terms of their numeric values. We explore the relationship between graph structure and the distribution of attribute values and propose an outlier-detection step, which is used as a constraint during substructure discovery. By pruning anomalous vertices and edges, more weight is given to the most descriptive substructures. Our method is applicable to multi-dimensional numeric attributes; we outline how it can be extended for high-dimensional data. We support our findings with experiments on transaction graphs and single large graphs from the domains of physical building security and digital forensics, measuring the effect on runtime, memory requirements and coverage of discovered patterns, relative to the unconstrained approach.

The Violent Maintenance of the Portuguese Colonial Identity and the Search for a Postcolonial one: literary images of Portugal as a colonial and postcolonial nation

Through the examination of Camões's Os Lusíadas , Sena's Os Grão-Capitães and Saramago's A Jangada de Pedra , this article explores violence as a means of shaping Portuguese identity in different historical contexts, and how these works portray the continued recourse to violence as Portugal moves from colonizing to postcolonial nation.

A search for star formation around the Galactic halo B-type star PHL 346 with the 2dF spectrograph

A preliminary search for stars that may have formed coevally with the apparently young halo B-type star PHL 346 has been performed with the 2dF multifibre spectrograph on the Anglo- Australian Telescope (AAT). Candidates were selected for spectroscopy from APM scans of B and R Schmidt plates centred on PHL 346. A total of 476 stars of spectral type A or F were found; radial velocity estimates and more accurate spectral type assignments narrowed the number of possible coeval candidates to 6 A-type and 14 F-type stars. A statistical analysis of these results using a comparison with a control field suggests that the number of A-type or F-type candidate stars around PHL 346 is not unexpected, and that they need not be associated with PHL 346. A number of ways to improve the project are suggested.

A search for circumstellar material around B-type stars in the Galactic halo

19 B-type stars, selected from the Palomar-Green Survey, have been observed at infrared wavelengths to search for possible infrared excesses, as part of an ongoing programme to investigate the nature of early-type stars at high Galactic latitudes. The resulting infrared fluxes, along with Stromgren photometry, are compared with theoretical flux profiles to determine whether any of the targets show evidence of circumstellar material, which may be indicative of post-main- sequence evolution. Eighteen of the targets have flux distributions in good agreement with theoretical predictions. However, one star, PG 2120 + 062, shows a small near-infrared excess, which may be due either to a cool companion of spectral type F5-F7, or to circumstellar material, indicating that it may be an evolved object such as a post-asymptotic giant branch star, in the transition region between the asymptotic giant branch and planetary nebula phase, with the infrared excess due to recent mass loss during giant branch evolution.

Search for 1-10 Hz modulations in coronal emission with SECIS during the August 11, 1999 eclipse

Results of the search of the periodic changes of the 530.3 nm line intensity emitted by selected structures of the solar corona in the frequency range 1-10 Hz are presented. A set of 12 728 images of the section of the solar corona extending from near the north pole to the south-west were taken simultaneously in the 530.3 nm ("green") line and white-light with the Solar Eclipse Coronal Imaging System (SECIS) during the 143-seconds- long totality of the 1999 August 11 solar eclipse observed in Shabla, Bulgaria. The time resolution of the collected data is better than 0.05 s and the pixel size is approximately 4 arcsec. Using classical Fourier spectral analysis tools, we investigated temporal changes of the local 530.3 nm coronal line brightness in the frequency range 1-10 Hz of thousands of points within the field of view. The various photometric and instrumental effects have been extensively considered. We did not find any indisputable, statistically significant evidence of periodicities in any of the investigated points (at significance level alpha = 0.05).

Description and search labor for information retrieval

Selection power is taken as the fundamental value for information retrieval systems. Selection power is regarded as produced by selection labor, which itself separates historically into description and search labor. As forms of mental labor, description and search labor participate in the conditions for labor and for mental labor. Concepts and distinctions applicable to physical and mental labor are indicated, introducing the necessity of labor for survival, the idea of technology as a human construction, and the possibility of the transfer of human labor to technology. Distinctions specific to mental labor, particular between semantic and syntactic labor, are introduced. Description labor is exemplified by cataloging, classification, and database description, can be more formally understood as the labor involved in the transformation of objects for description into searchable descriptions, and is also understood to include interpretation. The costs of description labor are discussed. Search labor is conceived as the labor expended in searching systems. For both description and search labor, there has been a progressive reduction in direct human labor, with its syntactic aspects transferred to technology, effectively compelled by the high relative costs of direct human labor compared to machine processes.