Search results

In mobile service domain, there are special problems needed to be considered: in the beginning of new mobile service running period, the number of customer is lower; During the running period, new customers continually join and/or service items are continually included into the service, so one customer often experience and evaluate limited items, which a can be found in many service applications such as web content recommendation.
For a specific Time Window T, the integration of all user’s context preferences for service s produces a context preference matrix RT, With the refinement of context grain size and the increase of service running time, RT will shows a high-dimensional characteristic and increasing capacity which means weak associations between customers appears.
Definition 3：is a CPH, for , the vertex weight and do not change, for,the initial hyperedge weight ,,k is the number of vertices in.For,,can be merged into one hyper edge ,.
GroupLens Organization provides a mobile version of Movielens website, mobile customer can browse and evaluate service offered by Movielens anywhere, anytime[9]. s is defined as a service which recommends comedy movie information, so the sample is chosen from dataset of comedy films.Cs = (gender, occupation, age,day time (morning, afternoon, evening), week (workday, weekend)),The value of last two context attributes are obtained by time stamp conversion.is the ratio of number of records with satisfactory feedback into the number of all records with satisfactory feedback.

As a result the wear of such multilayer coating would proceed in number of steps corresponding to number of layers and in the effect minimizing the substrate barred area, i.e. the overall wear.
The start of the wear both enlarge and increase number of these vertical cracks as well as opens some inclined to the surface through-crystalline.
A more detailed investigation of the wear grove center part and especially of the parts of the specimen of smaller thickness proved that TiN layers contain number of separate or multiple vertical cracks, while Ti layers are Fig.1.
This observation indicates that the fine graining together with some contamination from argon and nitrogen from deposition of Ti layer increased its strength above that of stainless steel substrate.

Along with the rapid population growth, industrial production scale expands unceasingly, the rapid development of urbanization, agricultural production with a large number of chemical fertilizers and sewage irrigation, and a lot of harmful substances into the soil system, the soil heavy metal pollution has become a serious environmental problem to the world.
The solid phase contains a large number of clay minerals, organic matter and metal oxide which can hold into its internal various pollutants, especially heavy metal elements, and then accumulate in the soil.
Heavy metal in edible part of crop and its effect on human health have a large number of reports.
A large number of heavy metal distributions in the root.
The heavy metal concentrations in the grain are relatively low.

Modelling the spike formation The Occam’s razor rule suggests we should interpret our experimental results with a minimal number of hypothesis.
It is known that solute interaction energies with grain boundaries are well correlated to the size effect of the solute atom.
This allows a ranking of the solute atoms in terms of increasing binding energies to the grain boundaries and, in the absence of any more detailed information, it is expected that the same classification (though with different binding energies) should hold for interactions with the interphase boundary.

The listing data suggests the necessity of a large number of parameters.
n –number of crystallization nuclei calculated for surface or volume unit, 1/m2 or 1/m3, DT – calculated local undercooling (from F-K model), [K] The parameters of above empirical equation (soft modeling): DTm-(i) – average undercooling corresponding to the maximum nucleation intensity, i=s (surface nucleation on casting-mould interface) or i=v (volume nucleation), 5,5 [K] DTσ – standard deviation, 0,5 [K] nmax – theoretical maximal nucleation density which can be reached when whole nucleation sites will be activated during cooling (nmax is determined experimentally for the given alloy), 2e10 [1/m3] The empirical grow model υ=f(t) (e.g.
Then the next stage of validation is presented, Principal chosen parameters governing the heat transfer, nucleation and growth phenomena: Hcast-insul (heat transf.coeff.)=10000 W/m2K Hcast-chilll (heat transf.coeff.)=2500 W/m2K 1 - Empirical nucleation model – nucleation distribution (Gauss distrib.), table 1 ΔTm-s(mean surf.undercooling) = 5K and ΔTm-s-Ch-HI=10 ns (max nuclei surf.number) = 1e5 1/m2 and ns-Ch = 8e5 ΔTm-v (mean volume undercooling)=2K nv (max nuclei volume number)=8e6 1/m3 σΔT=0,4 K 2 – Empirical dendrite growth model (Kurz-Giovanola-Trivedi), table 1] a2 (growth coeff.) =2,9e-3 mm/(sK-2) a3 (growth coeff.) =1e-9 mm/(sK-3) Oriented columnar structure Dendrite Arm Spacing (DAS) = 15 to25 μm (CET=Columnar Equiaxed )transition) 10mm 100μm CET experim= 62±6 CET simulat = 61±3 Cu – chill diam. 70 mm Cu – chill diam. 70 mm High insulating material High insulating material a.
Thevoz, A Three-Dimensional Cellular Automaton-Finite Element Model for the Prediction of Solidification Grain Structures, Metallurgical and Materials Transactions A, 1999, vol. 30A, pp. 3153-3165
-A., Rappaz M., A Coupled Finite ElementCellular Automaton Model for the Prediction of Dendritic Grain Structures in Solidification Processes, Acta Metall.

The information store the random number and ant’s 8 direction pheromone around order by column.
The random number also has the function of a flag to indicate whether the ant needs the next location information.
Algorithm data structure Every GPU thread based on the thread number corresponds to an ant space.
Threads based on the random number determine the next location of the moving and write to the corresponding positions in Result matrix.
A fine-grained parallel GPU accelerated ant colony algorithm [J].

A large number of micro-glass beads in it can reduce the friction between cement particles as ball bearing.
Thus the density difference between components of mixture is reduced and the sedimentation rate of grain will be greatly slowed down.
Table4 Mix proportions of lightweight aggregate concrete kg No Cement Fly ash Expanded perlite Expanding agent Water reducing agent W/C A1 400 450 150 40 8.5 0.7 A2 350 500 150 35 8.5 0.6 A3 300 550 150 30 8.5 0.5 A4 300 550 150 85 8.5 0.5 A5 300 550 150 85 8.5 0.55 Table5 Properties of lightweight aggregate concrete No Initial /h Final /h Bulk density /(g/cm3) viscosity Segregation Workability Bleeding Compressive Strength/MPa A1 2.9 9.6 1.32 better light Common light - A2 3.6 11.0 1.33 better light Common light - A3 4.5 10.4 1.31 best light better light 6.5 A4 4.1 12.1 1.34 best better Common no 6.6 A5 5.2 13.2 1.33 best better better no 6.9 Adding air entraining admixture and prewetted treatment.The proper amount of air entraining admixture can introduce a large number of tiny closed bubbles distributing evenly and stably in concrete structures gap.
Fig.1 shows pores are few and uneven, mostly penetrating in the reference of light aggregate concrete while a large number of unconnected spherical pores distribute evenly in this material, thus the bleeding channels are reduced and the collapse and segregation are effectively controlled in concrete.
(2)The addition of air entraining admixture can introduce a large number of tiny closed bubbles distributing evenly and stably in concrete structures gap, which reduces bleeding channels and the collapse causing by bleeding

We assume that the system has some NameNodes whose number is p.
The basic unit of metadata hash is DHU, so it is a coarse-grained mapping.
The maximum number of metadata in the client cache was set to 1500.
The number of clients was changed to test the performance of the system under different loads.
In the condition of single NameNode, the more of clients’ number was, the higher the average latency would be, and the average latency of the improved scheme was far lower than in other two conditions.

A number of functional nanofibers and composites have been developed through electrospinning by incorporating well-selected functional agents to achieve antibacterial properties.
FA is also found esterified to phytosterols present in grain products such as rice bran.
Antibacterial Activity.Antibacterial efficacy of surface modified nonwoven containing FA was investigated by estimating the number of viable bacteria cells in the S. aureus and E. coli suspension after being in contact with nonwoven for various periods of time, and the results were shown in Fig. 3.
Viable cell numbers of Staphylococcus aureus and Escherichia coli as a function time in contact with: nonwoven.
The cell number was determined by surface spread plate method.

Each copper nodule is made up by a large number of crystallites.
TEM micrograph shows a large number of crystallites with size ranging from 30-70 nm.
Fine crystallites indicate that a large number of atoms are trapped in the crystallite boundaries.
Higher energy zones like grain boundaries and twin boundaries [7] helps rapid biocidal effect due to the fast releasing of copper ions.
Each nodule is made up of a number of fine crystallites (size~50 nm).