Detecting digital chameleons

Conversations are characterized by an interactional synchrony between verbal and nonverbal behaviors [Kendon, A. (1970). Movement coordination in social interaction: some examples described. Acta Psychologica, 32(2), 101–125]. A subset of these contingent conversational behaviors is direct mimicry. During face to face interaction, people who mimic the verbal [Giles, H., Coupland, J., & Coupland, N. (1991). Accommodation theory: Communication, context, and consequence. In Giles, H., Coupland, J., & Coupland, N. Contexts of accommodation. Developments in applied sociolinguistics. Cambridge: Cambridge University Press] and nonverbal behaviors [Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: the perception-behavior link and social interaction. Journal of Personality and Social Psychology, 76, 893–910] gain social advantage. Most research examining mimicry behavior in interaction examines ‘implicit mimicry’ in which the mimicked individual is unaware of the behavior of the mimicker. In this paper, we examined how effective people were at explicitly detecting mimicking computer agents and the consequences of mimic detection in terms of social influence and interactional synchrony. In Experiment 1, participant pairs engaged in a “one-degree of freedom” Turing Test. When the computer agent mimicked them, users were significantly worse than chance at identifying the other human. In Experiment 2, participants were more likely to detect mimicry in an agent that mirror-mimicked their head movements (three degrees of freedom) than agents that either congruently mimicked their behaviors or mimicked those movements on another rotational axis. We discuss implications for theories of interactivity.     

Cloud-based video analytics using convolutional neural networks

Object classification is a vital part of any video analytics system, which could aid in complex applications such as object monitoring and management. Traditional video analytics systems work on shallow networks and are unable to harness the power of distributed processing for training and inference. We propose a cloud-based video analytics system based on an optimally tuned convolutional neural network to classify objects from video streams. The tuning of convolutional neural network is empowered by in-memory distributed computing. The object classification is performed by comparing the target object with the prestored trained patterns, generating a set of matching scores. The matching scores greater than an empirically determined threshold reveal the classification of the target object. The proposed system proved to be robust to classification errors with an accuracy and precision of 97% and 96%, respectively, and can be used as a general-purpose video analytics system.     

Complementary Channel Use and the Role of Social Competence

This study focused on channel complementarity among various interpersonal communication channels (face‐to‐face, telephone, e‐mail, text messaging, and Facebook). We looked at daily channel use among 136 participants and demonstrated complementary channel use among most combinations of channels, excluding face‐to‐face. We also extended channel complementarity theory by examining social competence as a moderator of channel complementarity. Results indicated that telephone and text messaging exhibited complementarity at high but not low levels of social competence, whereas e‐mail and text messaging exhibited complementarity at low but not high levels of social competence. Face‐to‐face communication and Facebook exhibited a displacement relationship at high but not low levels of social competence. Implications for channel complementarity theory and the role of individual characteristics are discussed.     

A more realistic approach for airport ground movement optimisation with stand holding

In addition to having to handle constantly increasing numbers of aircraft, modern airports also have to address a wide range of environmental regulations and requirements. As airports work closer and closer to their maximal possible capacity, the operations problems that need to be solved become more and more complex. This increasing level of complexity leads to a situation where the introduction of advanced decision support systems becomes more and more attractive. Such systems have the potential to improve efficient airside operations and to mitigate against the environmental impact of those operations. This paper addresses the problem of moving aircraft from one location within an airport to another as efficiently as possible in terms of time and fuel spent. The problem is often called the ground movement problem and the movements are usually from gate/stands to a runway or vice-versa. We introduce a new sequential graph based algorithm to address this problem. This approach has several advantages over previous approaches. It increases the realism of the modelling and it draws upon a recent methodology to more accurately estimate taxi times. The algorithm aims to absorb as much waiting time for delay as possible at the stand (with engines off) rather than out on the taxiways (with engines running). The impact of successfully achieving this aim is to reduce the environmental pollution. This approach has been tested using data from a European hub airport and it has demonstrated very promising results. We compare the performance of the algorithm against a lower bound on the taxi time and the limits to the amount of waiting time that can be absorbed at stand.     

Calculation of the Thermal Conductivity and Gas Permeability in a Uniaxial Bundle of Fibers

A model of the local microstructure of a bundle of fibers is simulated and used as the basis for calculations of transport properties. This, in turn, can be used in a macroscopic model of the chemical vapor infiltration process. An expanding/overlapping circle representation of the micro-structure simulates the deposition of matrix in a uniaxial bundle of fibers. An iterative heat conduction algorithm is used to calculate the transverse thermal conductivity based on the thermal conductivities of the solid and gas phases. The permeability of gas through the microstructure is calculated for flow both parallel and transverse to overlapping cylinders using a Stokes equation and assuming a Darcy's law behavior. Both the simulations of the microstructure and associated calculations of the transport properties compare favorably with experimental data. Darcy's law for the behavior of gas in a bundle of fibers is shown to be valid for gas pressures of 5–13 kPa.     

Sophorolipid-Derived Unsaturated and Epoxy Fatty Acid Estolides as Plasticizers for Poly(3-Hydroxybutyrate)

Unsaturated and epoxy fatty acid estolides were synthesized from the ω and ω‐1 hydroxy fatty acids derived from sophorolipids (SL). These estolides were utilized as additives in solution‐cast poly(3‐hydroxybutyrate) (P3HB) films and their plasticizing effects reported. Estolides in the P3HB film matrix resulted in slight reductions of melting and glass transition temperatures while the crystallinity remained relatively constant (±2.0 %). Scanning electron microscopy revealed irregular film surfaces in the presence of both estolides and the formation of pores within the P3HB film matrix. These irregularities influenced the tensile properties of the films by incrementally decreasing the tensile strength and moduli of the P3HB films and increasing their elongation. Curing the films for 3 months in the presence of estolide triggered an enhanced tensile strength and modulus when compared to the initial films. Irradiation was implemented in an attempt to crosslink the estolides. Results showed that radiation did not result in crosslinking but rather instigated chain scission and reduced the molecular weights by up to 81 % which further reduced the tensile strength, elongation and modulus of the P3HB films. These tensile property variations showed that while the tensile strength of the P3HB films declined in the presence of both estolides, the plasticity and elongation improved validating that these unique SL‐derived estolides can be effectively used as plasticizers in P3HB films.     

A peer-to-peer system for on-demand sharing of capacity across network applications

As a plethora of various distributed applications emerge, new computing platforms are necessary to support their extra and sometimes evolving requirements. This research derives its motive from deficiencies of real networked applications deployed on platforms unable to fully support their characteristics and proposes a network architecture to address that issue. Hoverlay is a system that enables logical movement of nodes from one network to another aiming to relieve requesting nodes, which experience high workload. Node migration and dynamic server overlay differentiate Hoverlay from Condor-based architectures, which exhibit more static links between managers and nodes. In this paper, we present a number of important extensions to the basic Hoverlay architecture, which collectively enhance the degree of control owners have over their nodes and the overall level of cooperation among servers. Furthermore, we carried out extensive simulations, which proved that Hoverlay outperforms Condor and Flock of Condors in both success rate and average successful query path length at a negligible increase in messages.     

Finite-Element Simulations of Full-Scale Modular-Block Reinforced Soil Retaining Walls under Earthquake Loading

A finite-element procedure was used to simulate the dynamic behavior of four full-scale reinforced soil retaining walls subjected to earthquake loading. The experiments were conducted at a maximum horizontal acceleration of over 0.8 g, with two walls subjected to only horizontal accelerations and two other walls under simultaneous horizontal and vertical accelerations. The analyzes were conducted using advanced soil and geosynthetic models that were capable of simulating behavior under both monotonic and cyclic loadings. The soil behavior was modeled using a unified general plasticity model, which was developed based on the critical state concept and that considered the stress level effects over a wide range of densities using a single set of parameters. The geosynthetic model was based on the bounding surface concept and it considered the S-shape load-strain behavior of polymeric geogrids. In this paper, the calibrations of the models and details of finite-element analysis are presented. The time response of horizontal and vertical accelerations obtained from the analyses, as well as wall deformations and tensile force in geogrids, were compared with the experimental results. The comparisons showed that the finite-element results rendered satisfactory agreement with the shake table test results.     

Using contests to allocate pollution rights

In this paper we advocate a new initial allocation mechanism for a tradable pollution permit market. We outline a Permit Allocation Contest (PAC) that distributes permits to firms based on their rank relative to other firms. This ranking is achieved by ordering firms based on an observable ‘external action’ where the external action is an activity or characteristic of the firm that is independent of their choice of emissions in the tradable permit market. We argue that this mechanism has a number of benefits over auctioning and grandfathering. Using this mechanism efficiently distributes permits, allows for the attainment of a secondary policy objective and has the potential to be more politically appealing than existing alternatives.