Search

Forecasting the popularity of new songs has become a standard practice in the music industry and provides a comparative advantage for those that do it well. Considerable efforts were put into machine learning prediction models for that purpose. It is known that in these models, relevant predictive parameters include intrinsic lyrical and acoustic characteristics, extrinsic factors (e.g., publisher influence and support), and the previous popularity of the artists. Much less attention was given to the social components of the spreading of song popularity. Recently, evidence for musical homophily - the tendency that people who are socially linked also share musical tastes - was reported. Here we determine how musical homophily can be used to predict song popularity. The study is based on an extensive dataset from the last.fm online music platform from which we can extract social links between listeners and their listening patterns. To quantify the importance of networks in the spreading of songs that eventually determines their popularity, we use musical homophily to design a predictive influence parameter and show that its inclusion in state-of-the-art machine learning models enhances predictions of song popularity. The influence parameter improves the prediction precision (TP/(TP+FN)) by about 50% from 0.14 to 0.21, indicating that the social component in the spreading of music plays at least as significant a role as the artist's popularity or the impact of the genre., Comment: 7 pages, 3 figures

Rediscovering a new function for something can be just as important as the discovery itself. In 1982, Stephen Jay Gould and Elisabeth Vrba named this phenomenon Exaptation to describe a radical shift in the function of a specific trait during biological evolution. While exaptation is thought to be a fundamental mechanism for generating adaptive innovations, diversity, and sophisticated features, relatively little effort has been made to quantify exaptation outside the topic of biological evolution. We think that this concept provides a useful framework for characterising the emergence of innovations in science. This article explores the notion that exaptation arises from the usage of scientific ideas in domains other than the area that they were originally applied to. In particular, we adopt a normalised entropy and an inverse participation ratio as observables that reveal and quantify the concept of exaptation. We identify distinctive patterns of exaptation and expose specific examples of papers that display those patterns. Our approach represents a first step towards the quantification of exaptation phenomena in the context of scientific evolution., Comment: 14 pages, 6 figures

The ImagingRing��� System (medPhoton, Salzburg, Austria) is a novel X-ray planar and cone beam computed tomography system for in room imaging in particle therapy. The aim of this study was to establish a Monte Carlo model of the ImagingRing��� System for future research on scatter effects. The X-ray head was modeled using the Monte Carlo toolkit GATE (v8.0, GEANT4 Application for Tomographic Emission) and GEANT4 (GEometry ANd Tracking v.10.3).In a first step, the tungsten anode and the electron beam emerging from the cathode were modeled using GATE. Its surrounding glass and oil, as well as the polycarbonate exit cone were modeled directly while the primary collimator, collimator jaws and flattening filter were imported from vendor supplied CAD-files. Next, experimental characterization was performed. Half-Value-Layers (HVL) in aluminum were determined with a selection of filtration levels and types using a NOMEX multimeter (PTW, Freiburg, Germany). In the simulated model, the energy spectrum was tuned by approximating the energy of the electron beam by a linear combination of discrete energies. The resulting MC based HVLs in aluminum were compared to experimental data. The best approximation for the energy spectrum was determined by minimizing a cost function. The physical dimensions of the electron focal spot on the anode were measured using a dedicated slit camera (PTW, Freiburg, Germany). The slit camera restricted the beam to a narrow slit. The projection of this slit was then registered by the detector of the ImagingRing��� System. Subsequently, findings on position and shape of the electron focal spot were implemented in the simulation. In addition, two-dimensional dose distributions in abscence of a flattening filter were first measured, using the scintillation based Lynx detector (IBA, Schwarzenbruck, Germany), and then simulated. The average deviation between measured and simulated HVLAl was within 3% for the whole clinical energy range between 80 keV to 120 keV. This agreement was X Iwithin measurement uncertainty. The size and shape of the simulated electron focal spot agreed closely with the experimentally measured data. The heel e_ect stemming from the anode was clearly visible in both the experimental data and the MC simulation and the intensity profiles matched. In conclusion, a GATE based X-ray head model was established, that accurately resembles experimental measurements. The presented method was shown to provide a realistic X-ray distribution, enabling the estimation of imaging doses when implementing new clinical protocols or predict the impact of technical changes of the X-ray source. The developed method can be transferred to model other commercial X-ray units. The established model also allows detailed MC based investigations of the head scatter to improve imaging quality.