A semantic approach to remove incoherent items from a user profile and improve the accuracy of a recommender system

Recommender systems usually suggest items by exploiting all the previous interactions of the users with a system (e.g., in order to decide the movies to recommend to a user, all the movies she previously purchased are considered). This canonical approach sometimes could lead to wrong results due to several factors, such as a change in user preferences over time, or the use of her account by third parties. This kind of incoherence in the user profiles defines a lower bound on the error the recommender systems may achieve when they generate suggestions for a user, an aspect known in literature as magic barrier. This paper proposes a novel dynamic coherence-based approach to define the user profile used in the recommendation process. The main aim is to identify and remove, from the previously evaluated items, those not semantically adherent to the others, in order to make a user profile as close as possible to the user’s real preferences, solving the aforementioned problems. Moreover, reshaping the user profile in such a way leads to great advantages in terms of computational complexity, since the number of items considered during the recommendation process is highly reduced. The performed experiments show the effectiveness of our approach to remove the incoherent items from a user profile, increasing the recommendation accuracy.     

Interplay between upsampling and regularization for provider fairness in recommender systems

User Modeling and User-Adapted Interaction - Considering the impact of recommendations on item providers is one of the duties of multi-sided recommender systems. Item providers are key stakeholders...     

Subdynamics as a mechanism for objective description

Abstract

The relationship between microsystems and macrosystems is considered in the context of quantum field formulation of statistical mechanics: it is argued that problems on foundations of quantum mechanics can be solved relying on this relationship. This discussion requires some improvement of non-equilibrium statistical mechanics which is briefly presented.     

Reforming of olive mill wastewater through a Pd-membrane reactor

This work investigates the exploitation of olive mill wastewater (OMW) for producing hydrogen in a membrane reformer.     

Creating an Holistic Emergency Alert Management Platform

ABSTRACT

Extreme natural events require effective emergency procedures to minimize adverse effects on a region’s population and economy. Such procedures typically involve the effort of several different teams of first responders (e.g., fire fighters, public administrations, police departments, utility companies), hence coordination is fundamental to the effectiveness of the response to the emergency that must be supported with adequate infrastructures. Nonetheless, first responders often rely on manual processes, in the life cycle of extreme events, which do not change consistently with the type of shock or affected population. The aim of this paper is to present a technology transfer process to improve both the emergency alert process and the knowledge of disaster-type safety procedures through the implementation of a proposed platform. We also highlight a pilot application on a post-disaster case study—the province of L’Aquila (Abruzzi) in Italy.     

Large-Area Direct Hetero-Epitaxial Growth of 1550-nm InGaAsP Multi-Quantum-Well Structures on Patterned Exact-Oriented (001) Silicon Substrates by Metal Organic Chemical Vapor Deposition

We employ a simple two-step growth technique to grow large-area 1550-nm laser structures by direct hetero-epitaxy of III–V compounds on patterned exact-oriented (001) silicon (Si) substrates by metal organic chemical vapor deposition. Densely-packed, highly uniform, flat and millimeter-long indium phosphide (InP) nanowires were grown from Si v-grooves separated by silicon dioxide (SiO₂) stripes with various widths and pitches. Following removal of the SiO₂ patterns, the InP nanowires were coalesced and, subsequently, 1550-nm laser structures were grown in a single overgrowth without performing any polishing for planarization. X-ray diffraction, photoluminescence, atomic force microscopy and transmission electron microscopy analyses were used to characterize the epitaxial material. PIN diodes were fabricated and diode-rectifying behavior was observed.     

Three-dimensional silicon inverse photonic quasicrystals for infrared wavelengths

Quasicrystals are a class of lattices characterized by a lack of translational symmetry. Nevertheless, the points of the lattice are deterministically arranged, obeying rotational symmetry. Thus, we expect properties that are different from both crystals and glasses. Indeed, naturally occurring electronic quasicrystals (for example, AlPdMn metal alloys) show peculiar electronic, vibrational and physico-chemical properties. Regarding artificial quasicrystals for electromagnetic waves, three-dimensional (3D) structures have recently been realized at GHz frequencies and 2D structures have been reported for the near-infrared region. Here, we report on the first fabrication and characterization of 3D quasicrystals for infrared frequencies. Using direct laser writing combined with a silicon inversion procedure, we achieve high-quality silicon inverse icosahedral structures. Both polymeric and silicon quasicrystals are characterized by means of electron microscopy and visible-light Laue diffraction. The diffraction patterns of structures with a local five-fold real-space symmetry axis reveal a ten-fold symmetry as required by theory for 3D structures.     

On the nature and importance of the transition between molecules and nanocrystals: towards a chemistry of "nanoscale perfection"

This paper discusses the importance of the transition between molecular compounds and nanocrystals. The boundary between molecular and nanocrystals/nanoclusters can be defined by the emergence of the bulk phase; atoms in the core of the nanoclusters that are not bound to ligands. This transition in dimensions and structural organization is important because it overlaps with the boundary between atomically defined moieties (molecules can be isolated with increasing purity) and mixtures (nanocrystals have a distribution of sizes, shapes, and defects; they cannot be easily separated into batches of structurally identical species). Passing through this boundary, as the size of a structure increases beyond a few nanometres, the information about the position of each atom gradually disappears. This loss of structural information about a chemical structure fundamentally compromises our ability to use it as a part of a complex chemical system. If we are to engineer complex functions encoded in a chemical language, we will need pure batches of atomically defined (truly monodisperse) nanoscale compounds, and we will need to understand how to make them and preserve them over a broad range of length scales, compositions, and timeframes. In this review we survey most classes of monodisperse nanomaterials (mostly nanoclusters) and highlight the recent breakthroughs in this area which might be spearheading the development of a chemistry of "nanoscale perfection".