Exploring latent browsing graph for question answering recommendation

In this paper, we develop a framework of Question Answering Pages (referred to as QA pages) recommendation. Our proposed framework consists of the two modules: the off-line module to determine the importance of QA pages and the on-line module for on-line QA page recommendation. In the off-line module, we claim that the importance of QA pages could be discovered from user click streams. If the QA pages are of higher importance, many users will click and spend their time on these QA pages. Moreover, the relevant relationships among QA pages are captured by the browsing behavior on these QA pages. As such, we exploit user click streams to model the browsing behavior among QA pages as QA browsing graph structures. The importance of QA pages is derived from our proposed QA browsing graph structures. However, we observe that the QA browsing graph is sparse and that most of the QA pages do not link to other QA pages. This is referred to as a sparsity problem. To overcome this problem, we utilize the latent browsing relations among QA pages to build a QA Latent Browsing Graph. In light of QA latent browsing graph, the importance score of QA pages (referred to as Latent Browsing Rank) and the relevance score of QA pages (referred to as Latent Browsing Recommendation Rank) are proposed. These scores demonstrate the use of a QA latent browsing graph not only to determine the importance of QA pages but also to recommend QA pages. We conducted extensive empirical experiments on Yahoo! Asia Knowledge Plus to evaluate our proposed framework.     

Faster Approximate String Matching for Short Patterns

We study the classical approximate string matching problem, that is, given strings P and Q and an error threshold k, find all ending positions of substrings of Q whose edit distance to P is at most k. Let P and Q have lengths m and n, respectively. On a standard unit-cost word RAM with word size w≥log n we present an algorithm using time

On the network effect in Web 2.0 applications

In recent years, the proliferation of the world wide web has lead to an increase in a number of applications such as search, social networks and auctions, whose success depends critically upon the number of users of that service. Two examples of such applications are internet auctions and social networks. One of the characteristics of online auctions is that a successful implementation requires a high volume of buyers and sellers at its website. Consequently, auction sites which have a high volume of traffic have an advantage over those in which the volume is limited. This results in even greater polarization of buyers and sellers towards a particular site. The same is true for social networks in which greater use of a given social network increases the use from other participants on the network. This is often referred to as the “network effect” in a variety of interaction-centric applications in networks. While this effect has qualitatively been known to increase the value of the overall network, its effect has never been modeled or studied rigorously. In this paper, we construct a Markov model to analyze the network effect in the case of two important classes of web applications. These correspond to auctions and social networks. We show that the network effect is very powerful and can result in a situation in which an auction or a social networking site can quickly overwhelm its competing sites. Thus, the results of this paper show the tremendous power of the network effect for Web 2.0 applications.     

Advances in Applying Computer-Aided Drug Design for Neurodegenerative Diseases

Neurodegenerative diseases (NDs) including Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis, and Huntington’s disease are incurable and affect millions of people worldwide. The development of treatments for this unmet clinical need is a major global research challenge. Computer-aided drug design (CADD) methods minimize the huge number of ligands that could be screened in biological assays, reducing the cost, time, and effort required to develop new drugs. In this review, we provide an introduction to CADD and examine the progress in applying CADD and other molecular docking studies to NDs. We provide an updated overview of potential therapeutic targets for various NDs and discuss some of the advantages and disadvantages of these tools.     

X-ray photoelectron spectroscopic study of the oxidation and reduction of a cerium(III) oxide/cerium foil substrate

X-ray photoelectron spectroscopy has been used to examine the nature of the oxide overlayers on a passivated cerium metal foil as a function of a variety of oxidation and reduction treatments. Oxidation of a clean uncontaminated cerium(III) oxide surface is facile at room temperature and produces non-stoichiometric ceria (CeO_2–x) at oxygen doses as low as 10 L. At higher doses the overlayer thickens, and after a dose of 160 L the layer depth exceeds the Ce 3d photoelectron attenuation distance of about 20 Å. High pressure treatment of the foil in oxygen (0.5 bar at RT and 473 K) produces CeO₂ in a high degree of crystallographic order such that O 1s photoelectron intensities are increased above that expected from a randomly oriented powder. An attempt to reduce the CeO₂ layer formed by controlled oxidation with CO (633 K, 14 h, 0.6 bar) results in the formation of a carbonated surface layer. Results following attempts to reoxidise this layer are discussed.     

Fatty acid selectivity of lipases: Erucic acid from rapeseed oil

The fatty acid selectivity of several commercial lipases was evaluated in the hydrolysis of high-erucic acid rapeseed oil (HEARO). The lipase ofPseudomonas cepacia catalyzed virtually complete hydrolysis of the oil (94–97%), while that ofGeotrichum candidum discriminated strongly against erucic acid, especially in esterification. A two-step process is suggested for obtaining a highly enriched erucic acid in which theG. candidum lipase is employed to selectively esterify the fatty acid residues of unsaturated C-18, and shorter chain acids, from a mixture of HEARO fatty acids obtained from total hydrolysis of the oil withP. cepacia lipase.     

A New Level-Set Model for the Representation of Non-Smooth Geometries

In Starinshak et al. (J Comput Phys 262(1):1–16, 2014), we proposed a new level-set model for representing multimaterial flows in multiple space dimensions. Rather than associating each level-set function with the boundary of a material, the new model associates each level-set function with a pair of materials and the interface that separates them. In this paper, we extend the model to represent geometries with non-smooth boundaries. The model uses multiple level-set functions to describe the shape boundary, typically with one level-set function per smooth boundary segment. Sign information is collected from all level-set functions and a voting algorithm is used to determine the interior/exterior of the geometric shape. The model is well suited for representing boundaries with singularities; it offers significant improvement over standard level-set approaches, both in shape preservation and area conservation; and it eliminates the need for costly redistancing of the level-set function. Numerical examples illustrate the superior performance of the proposed model.     

Bulk and suspension polymerization of styrene in the presence of n-pentane. An evaluation of monofunctional and bifunctional initiation

The plasticizing effect of n-pentane on the rate of bulk free radical polymerization of styrene and molecular weight distribution development has been modeled on the basis of the free volume theory for both monofunctional and bifunctional initiation. A strong decrease in the reaction rate in the late stages of the polymerization, due to the displacement of the onset of the gel effect, has been observed for both types of initiation. This decrease in the polymerization rate limited the terminal conversion to values well below 100% for mono-functional initiation. However, in bifunctionally initiated polymerization, terminal conversions close to 100% were obtained in spite of the decrease in reaction rate. Contrary to what was expected, the molecular weight distribution obtained at terminal conversion was almost completely insensitive to these changes in polymerization rate. This phenomenon is explained in terms of limited transfer to monomer reactions when n-pentane is present in the system. In suspension polymerization, the limiting conversion and plasticizing effects of n-pentane in monofunctionally initiated systems, caused enhanced coalescence leading to suspension set-up. In bifunctionally initiated systems this enhanced coalescence was completely overcome by the short duration of the particle growth stage, owing to high polymerization rates, and stable suspensions were achieved. For these systems the particle size distributions obtained were similar to that of suspension polystyrene without n-pentane. © 1993 John Wiley & Sons, Inc.