A Compositional and Interpretable Semantic Space

Vector Space Models (VSMs) of Semantics are useful tools for exploring the semantics of single words, and the composition of words to make phrasal meaning. While many methods can estimate the meaning (i.e. vector) of a phrase, few do so in an interpretable way. We introduce a new method (CNNSE) that allows word and phrase vectors to adapt to the notion of composition. Our method learns a VSM that is both tailored to support a chosen semantic composition operation, and whose resulting features have an intuitive interpretation. Interpretability allows for the exploration of phrasal semantics, which we leverage to analyze performance on a behavioral task.

Evidence of syntactic working memory usage in MEG data

While reading times are often used to measure working memory load, frequency effects (such as surprisal or n-gram frequencies) also have strong confounding effects on reading times. This work uses a naturalistic audio corpus with magnetoencephalographic (MEG) annotations to measure working memory load during sentence processing. Alpha oscillations in posterior regions of the brain have been found to correlate with working memory load in non-linguistic tasks (Jensen et al., 2002), and the present study extends these findings to working memory load caused by syntactic center embeddings. Moreover, this work finds that frequency effects in naturally-occurring stimuli do not significantly contribute to neural oscillations in any frequency band, which suggests that many modeling claims could be tested on this sort of data even without controlling for frequency effects.

Association of Functional Heme Oxygenase-1 Gene Promoter Polymorphism with Renal Transplantation Outcomes

Heme oxygenase-1 (HO-1) is a cytoprotective molecule and increased expression in experimental transplant models correlates with reduced graft injury. A functional dinucleotide repeat (GT)n polymorphism, within the HO-1 promoter, regulates gene expression; a short number of repeats (S-allele

The horse in early Ireland

The main object of this paper is to review the development of horse exploitation in Ireland between its introduction in the Bronze Age and the medieval period. The review considers the evidence of the use of horse for riding and traction and contrasts this with the evidence from neighbouring Britain. The change in horse size is traced as is the development of horse-related technology. The association of horse with burial ritual and the inauguration of kings is also considered.

Ion Association in [bmim][PF6]/Naphthalene Mixtures: An Experimental and Computational Study

Mixtures of room temperature ionic liquids (IL) with neutral organic molecules provide a valuable testing ground to investigate the interplay of the ionic and molecular-dipolar state in dense Coulomb systems at near ambient conditions. In the present study, the viscosity eta and the ionic conductivity a of 1-n-butyl-3-methylimidazolium hexafluorophosphate ([bmim][PF6])/naphthalene mixtures at T = 80 degrees C have been measured at 10 stoichiometries spanning the composition range from pure naphthalene to pure [bmim][PF6]. The viscosity grows nearly monotonically with increasing IL mole fraction (x), whereas the conductivity per ion displays a clear peak at x approximate to 15%. The origin of this maximum has been investigated using molecular dynamics simulations based on a classical force field. Snapshots of the simulated samples show that the conductivity maximum is due to the gradual transition in the IL component from an ionic state at high x to a dipolar fluid made of neutral ion pairs at low x. At concentrations x <0.20 the ion pairs condense into molecular-thin filaments bound by dipolar forces and extending in between nanometric droplets of IL. These results are confirmed and complemented by the computation of dynamic and transport properties in [bmim][PF6]/naphthalene mixtures at low IL concentration.

Insight into the adsorption competition and the relationship between dissociation and association reactions in ammonia synthesis

Ammonia synthesis on three metal surfaces (Zr, Ru, and Pd) is investigated using density functional theory calculations. In addition to N-2 dissociation, all the transition states of the hydrogenation reactions from N to NH3 are located and the reaction energy profiles at both low and high surface coverages are compared and analyzed. The following are found: (i) Surface coverage effect on dissociation reactions is more significant than that on association reactions. (ii) The difference between N and H chemisorption energies, the so-called chemisorption energy gap which is a measure of adsorption competition, is vital to the reactivity of the catalysts. (iii) The hydrogenation barriers can considerably affect the overall rate of ammonia synthesis. A simple model to describe the relationship between dissociation and association reactions is proposed. (c) 2007 American Institute of Physics.

Insight into association reactions on metal surfaces: Density-functional theory studies of hydrogenation reactions on Rh(111)

Hydrogenation reaction, as one of the simplest association reactions on surfaces, is of great importance both scientifically and technologically. They are essential steps in many industrial processes in heterogeneous catalysis, such as ammonia synthesis (N-2+3H(2)-->2NH(3)). Many issues in hydrogenation reactions remain largely elusive. In this work, the NHx (x=0,1,2) hydrogenation reactions (N+H-->NH, NH+H-->NH2 and NH2+H-->NH3) on Rh(111) are used as a model system to study the hydrogenation reactions on metal surfaces in general using density-functional theory. In addition, C and O hydrogenation (C+H-->CH and O+H-->OH) and several oxygenation reactions, i.e., C+O, N+O, O+O reactions, are also calculated in order to provide a further understanding of the barrier of association reactions. The reaction pathways and the barriers of all these reactions are determined and reported. For the C, N, NH, and O hydrogenation reactions, it is found that there is a linear relationship between the barrier and the valency of R (R=C, N, NH, and O). Detailed analyses are carried out to rationalize the barriers of the reactions, which shows that: (i) The interaction energy between two reactants in the transition state plays an important role in determining the trend in the barriers; (ii) there are two major components in the interaction energy: The bonding competition and the direct Pauli repulsion; and (iii) the Pauli repulsion effect is responsible for the linear valency-barrier trend in the C, N, NH, and O hydrogenation reactions. For the NH2+H reaction, which is different from other hydrogenation reactions studied, the energy cost of the NH2 activation from the IS to the TS is the main part of the barrier. The potential energy surface of the NH2 on metal surfaces is thus crucial to the barrier of NH2+H reaction. Three important factors that can affect the barrier of association reactions are generalized: (i) The bonding competition effect; (ii) the local charge densities of the reactants along the reaction direction; and (iii) the potential energy surface of the reactants on the surface. The lowest energy pathway for a surface association reaction should correspond to the one with the best compromise of these three factors. (C) 2003 American Institute of Physics.