Towards an efficient static scheduling scheme for delivering queries to heterogeneous clusters in the similarity search problem

Medium and large clusters incorporating hybrid CPU/graphics processing unit (GPU) nodes are present in many datacenters today. They can accelerate many different kinds of applications and appropriately manage applications dealing with a high volume of data. This is the case of the similarity problem because large databases are managed and very quick responses are required to hundreds or thousands of queries per second. However, the design and usage of heterogeneous computing platforms poses big challenges as system size, energy saving, task mapping, scheduling, among others, must be efficiently handled. In this paper we focus on the scheduling issue for distributing the incoming queries to all the processing components in the cluster nodes. Our algorithms exploit the computational resources, simultaneously processing queries on CPU cores and on the GPUs. Thus, we address the problem of how to distribute the queries over the whole system in order to obtain the best performance, under the assumption of defining a heuristic that automatically provides the best distribution. Experimental results show the benefits in terms of execution time and energy saving of using an appropriate scheduling scheme.     

Exchange of domains of glutamate dehydrogenase from the hyperthermophilic archaeon Pyrococcus furiosus and the mesophilic bacterium Clostridium difficile: effects on catalysis, thermoactivity and stability

The glutamate dehydrogenase gene from the hyperthermophilicarchaeon Pyrococcus furiosus has been functionally expressedin Escherichia coli under the control of the X, PL promoter.The P.furiosus glutamate dehydrogenase amounted to 20% of thetotal E.coli cell protein, and the vast majority consisted ofhexamers. Following activation by heat treatment, an enzymecould be purified from E.coli that was indistinguishable fromthe glutamate dehydrogenase purified from P.furiosus. Hybridgenes, that consisted of the coding regions for the homologousglutamate dehydrogenases from P.furiosus and the mesophilicbacterium Clostridium difficile, were constructed and successfullyexpressed in E.coli. One of the resulting hybrid proteins, containingthe glutamate binding domain of the C.difficile enzyme and thecofactor binding domain of the P.furiosus enzyme, did not showa detectable activity. In contrast, the complementary hybridcontaining the P.furiosus glutamate and the C.difficile cofactorbinding domain was a catalytically active hexamer that showeda reduced substrate affinity but maintained efficient cofactorbinding with the specificity found in the Clostridium symbiosumenzyme. Compared with the C.difficile glutamate dehydrogenase,the archaeal-bacterial hybrid is slightly more thermoactive,less thermostable but much more stable towards guanidinium chloride-inducedinactivation and denaturation     

Preparation of β-Cyclodextrin-Dextran Polymers and their Use as Supramolecular Carrier Systems for Naproxen

Summary Dextran, previously activated by periodate oxidation, was grafted with mono-6-amino-6-deoxy-β-cyclodextrin, mono-6-ethylenediamino-6-deoxy-β-cyclodextrin, and mono-6-butylenediamino-6-deoxy-β-cyclodextrin by reductive alkylation in the presence of NaBH₄. The polymers were able to form inclusion complexes with Naproxen, increasing the solubility of the drug by 2.2-2.6 folds. The β-cyclodextrin-grafted dextrans were used as macromolecular carriers for Naproxen, improving the “in vivo” anti-inflammatory activity of the drug.     

Calibration via Multi-period State Estimation in Water Distribution Systems

Calibration of model parameters is of utmost importance to ensure the good performance of hydraulic simulation models. In this work, calibration is conceived within a joint multi-period parameter and state estimation approach, where model parameters (i.e. roughness coefficients) and hydraulic variables should be computed from available measurements at different times. The aim of this paper is twofold: (1) to present a novel methodology for the calibration of water networks via multi-period state estimation, and (2) to adapt observability analysis to this approach. The novelty of this work is that such a large-scale non-linear optimisation problem is here solved using mathematical programming decomposition techniques. On the other hand, observability analysis requires the construction of the multi-period measurement and parameter Jacobian matrix of the problem. The proposed approach enables computation of the observable roughness coefficients from available readings over time, making possible the periodic reassessment of roughness values based on recent online measurements. The potential of the method is illustrated by means of a case study, which shows how such a methodology would contribute to make the most of telemetry data for calibration purposes.     

Reactive blending of commercial PET and PC with freshly added catalysts

Reactive blending of commercial poly(ethylene terephthalate) (PET) and bisphenol A polycarbonate (PC), with catalysts added in the form of powder dispersed on the polymers just before melt mixing, was performed in a Brabender Plasticord 2000 apparatus at 275°C. Catalytic activity of the catalysts freshly added to polymers was found to be much higher as compared with that of the residues of the same type of catalysts remaining in PET after its synthesis. Furthermore, the catalytic activity appeared to be strongly dependent on the structure of the ligand that influences the catalyst solubility in the polymer melt. N.m.r. spectroscopy, selective degradation of PC fragments, solubility tests in methylene chloride and d.s.c. measurements made it possible to range the catalysts studied according to their catalytic activity.     

Bulk functionalization of ethylene–propylene copolymers. II. Influence of the initiator concentration and of the copolymer composition and chain microstructure on the reaction kinetics

The effect of the dicumylperoxide (DCPO) concentration on the kinetics of the bulk functionalization of an ethylene-propylene copolymer (EPR) has been studied. It has been found that, enhancing the DCPO initial concentration in the reaction mixture, results in both increased initial reaction rate and grafting process efficiency. DCPO content also affects degradation, which always accompanies the grafting reaction. At high DCPO content, products with a high degree of grafting and intrinsic viscosity close to that of the parent copolymer can be obtained. Furthermore, the influence of the EPR composition and chain microstructure on the functionalization as well as on the degradation processes has been investigated. By increasing the C₂ content in EPRs prepared by using vanadium-based conventional catalysts, obtains faster functionalization kinetics and a less pronounced degradation of the products. At fixed experimental conditions, the highest grafting degrees have been obtained by using EPRs prepared by high-yield titanium catalysts and characterized as a block-like microstructure. The experimental results have been interpreted on the basis of a proposed reaction mechanism.     

Epoxy Networks Modified by a New Class of Oligomeric Silsesquioxanes Bearing Multiple Intramolecular Rings Formed through Si?O?C Bonds

Summary: A new class of silsesquioxane (SSO), containing species with two to nine Si atoms bearing multiple intramolecular rings formed through Si? O? C bonds, was synthesized as a glassy powder. It was characterized by UV‐MALDI‐TOF MS, ²⁹Si NMR and FT IR. Solutions containing different amounts of SSO in the diglycidyl ether of bisphenol A (DGEBA), were homopolymerized in the presence of (4‐dimethylamino)pyridine (DMAP) as initiator, leading to SSO‐modified epoxy networks. SSO species were covalently bonded to the epoxy network without any evidence of phase separation. The SSO addition provoked an increase in the elastic modulus in the glassy state explained by an increase in the cohesive energy density. The SSO addition gave also place to an increase in the intensity of tan δ and a decrease in both the glass transition temperature and the elastic modulus in the rubbery state. This was explained by a decrease in crosslink density associated with the flexibility of SSO structures. DMAP was much more effective than other usual initiators (like benzyldimethylamine, BDMA), in increasing the crosslink density of the resulting epoxy network. This led to high values of the glass transition temperature and the elastic modulus in the rubbery state.

Schematic representation of the chemical structure of the most significant species containing three Si atoms, present in the silsesquioxane.     

Windows Mobile LiveSD Forensics

More and more often, smartphones are relevant targets of civil and criminal investigations. Currently, there are several tools available to acquire forensic evidence from smartphones. Unfortunately, most of these tools require to connect the smartphone under investigation through a cable to an external device, like a computer or a multimeter. Some tools even require to disassemble the chips from the smartphone board.In this paper, we propose LiveSD Forensics, an on-device live data acquisition solution, to acquire evidence from both the Random-Access Memory (RAM) and the Electronically Erasable Programmable Read Only Memory (EEPROM) of Windows Mobile Devices.To the best of our knowledge, LiveSD Forensics is the only tool that performs on-device live data acquisition of the RAM and the EEPROM of Windows Mobile Devices. LiveSD Forensics uses a standard SD-Card equipped with tailored code to perform the data acquisition. Compared to other existing tools, LiveSD also generates the smallest memory alteration. Finally, to assess the effectiveness of the proposed methodology, we test LiveSD in a practical scenario, that is retrieving from the RAM the cryptographic key used by a known on-the-fly encryption tool. Results support the quality and effectiveness of our proposal.     

A reverse engineering approach for automatic annotation of Web pages

The Semantic Web is gaining increasing interest to fulfill the need of sharing, retrieving, and reusing information. Since Web pages are designed to be read by people, not machines, searching and reusing information on the Web is a difficult task without human participation. To this aim adding semantics (i.e meaning) to a Web page would help the machines to understand Web contents and better support the Web search process. One of the latest developments in this field is Google’s Rich Snippets, a service for Web site owners to add semantics to their Web pages. In this paper we provide a structured approach to automatically annotate a Web page with Rich Snippets RDFa tags. Exploiting a data reverse engineering method, combined with several heuristics, and a named entity recognition technique, our method is capable of recognizing and annotating a subset of Rich Snippets’ vocabulary, i.e., all the attributes of its Review concept, and the names of the Person and Organization concepts. We implemented tools and services and evaluated the accuracy of the approach on real E-commerce Web sites.