Validation of scientific topic models using graph analysis and corpus metadata

Probabilistic topic modeling algorithms like Latent Dirichlet Allocation (LDA) have become powerful tools for the analysis of large collections of documents (such as papers, projects, or funding applications) in science, technology an innovation (STI) policy design and monitoring. However, selecting an appropriate and stable topic model for a specific application (by adjusting the hyperparameters of the algorithm) is not a trivial problem. Common validation metrics like coherence or perplexity, which are focused on the quality of topics, are not a good fit in applications where the quality of the document similarity relations inferred from the topic model is especially relevant. Relying on graph analysis techniques, the aim of our work is to state a new methodology for the selection of hyperparameters which is specifically oriented to optimize the similarity metrics emanating from the topic model. In order to do this, we propose two graph metrics: the first measures the variability of the similarity graphs that result from different runs of the algorithm for a fixed value of the hyperparameters, while the second metric measures the alignment between the graph derived from the LDA model and another obtained using metadata available for the corresponding corpus. Through experiments on various corpora related to STI, it is shown that the proposed metrics provide relevant indicators to select the number of topics and build persistent topic models that are consistent with the metadata. Their use, which can be extended to other topic models beyond LDA, could facilitate the systematic adoption of this kind of techniques in STI policy analysis and design.

     

Influence of green creativity on organizations: A case study from the perspectives of leaders and subordinates

The sustainability framework has a strong influence on the policies, practices and procedures of companies. Green creativity plays a pivotal role in the development of initiatives and innovations for the environmental pillar of sustainability. Based on a survey applied to a Portuguese company with an accredited environmental management system (N = 146), the findings reveal that green creativity is influenced in different ways when analysed from distinct perspectives. The leader's green creativity is positively influenced by the subordinate's green self-efficacy and personal identification with the leader, whereas the subordinate's green creativity is positively influenced by their green passion and green self-efficacy, and the leaders' green creativity. This study shows that green creativity develops differentially according to distinct job positions and expectations that in practice must be complementary and synergistic for its effective development.     

An efficient blocking-matching algorithm based on fuzzy reasoning

Due to the temporal and spatial correlation of image sequence, the motion vector of a reference block is highly related to the motion vectors of its adjacent blocks in the same image frame. By using that idea, we propose a novel efficient fuzzy search (EFS) algorithm for block motion estimation. The experimental results show that the EFS performs better than other fast search algorithms, such as TSS, CS, NTSS, FSS, BBGDS, SES, and PSA in terms of picture quality, accuracy, computational complexity, and coding efficiency.     

Exercise training-induced changes in sensitivity to endothelin-1 and aortic and cerebellum lipid profile in rats

The purpose of this work was to study whether exercise training induces changes in the lipid profile of rat aorta and nervous system and in the in vitro intrinsic responsiveness of these tissues to endothelin-1 (ET-1) treatment. The exercise program performed successfully produced the characteristic metabolic alterations of the trained state. Exercise training induced a large and significant increase in the levels of both aortic ethanolamine plasmalogens (PlasEtn) and glucosylceramides. In contrast, a decrease of aortic ceramide and cholesterol levels was evoked by exercise training. ET-1 increased PlasEtn content only in sedentary animals. An exercise-induced increase in cerebellum levels of ceramides and ceramide monohexosides was found. The cerebellum ceramide content was increased by FT-1 more noticeably in sedentary rats than in trained animals. In contrast, cerebral cortex was observed to be largely insensitive to both exercise training and ET-1 treatment. It was concluded that exercise training (i) induces changes in both vascular and cerebellar lipid profiles, the former being much more pronounced than the latter, and (ii) diminishes the aortic and cerebellar sensitivity to ET-1 action.     

V2O5 nanofibre sheet actuators

Vanadium oxides, such as V2O5, are promising for lithium-ion batteries, catalysis, electrochromic devices and sensors. Vanadium oxides were proposed more than a decade ago for another redox-dependent application: the direct conversion of electrical energy to mechanical energy in actuators (artificial muscles). Although related conducting polymer and carbon nanotube actuators have been demonstrated, electromechanical actuators based on vanadium oxides have not be realized. V2O5 nanofibres and nanotubes provide the potential advantages of low-cost synthesis by sol-gel routes and high charging capacity and long cycle life. Here, we demonstrate electromechanical actuation for obtained high modulus V2O5 sheets comprising entangled V2O5 nanofibres. The high surface area of these V2O5 sheets facilitates electrochemical charge injection and intercalation that causes the electromechanical actuation. We show that the V2O5 sheets provide high Young's modulus, high actuator-generated stress, and high actuator stroke at low applied voltage.     

Design issues and performance analysis for DSR routing with reclaim‐based caching in MANETs

Mobile ad hoc networks (MANETs) have been positioned as one of the most important emerging wireless communication scenarios. Temporally formed by a collection of wireless mobile hosts, a MANET does not require the aid of any centralized administration. From this stems a suite of challenges in achieving an efficient MANET routing and content delivery in order to make the best use of precious resources and reduce the routing overhead at each MANET host. One of the reported approaches for solving the issue is the use of caching, which is expected to minimize the routing overhead by taking advantage of the limited memory at each mobile host. This paper introduces a novel scheme for addressing the above issue, called Reclaim‐Based Caching (RBC) policy, which dynamically utilizes a cache replacement mechanism of reclaiming stale routes to efficiently verify and validate a recoverable caching mechanism in Dynamic Source Routing. The main design purposes of RBC are to reduce the routing control overhead, lower the end‐to‐end routing delay, enhance the packet delivery ratio, and obtain a higher throughput for improving routing performance and accelerating the Route Discovery process due to low temporary link failure and high cache utilization. Hence, we can gather all feasible and historical route information into the cache library to be reclaimed as a threshold of efficient routing control. Consequently, the proposed RBC of this paper can be used in the universal wireless network environment to achieve better routing performance and to provide a more flexible real‐time application. Copyright © 2009 John Wiley & Sons, Ltd.     

An efficient VLSI switch-box router

A tool for switch-box routing that can route regions with cyclic constraints and with terminals on three or four sides is presented. A divide-and-conquer algorithm is used to explore the greedy channel routing idea, using techniques such as routing area partitioning, dynamic routing strategies, and sweeping concurrent bidirectional columns. The routing area is decomposed into three parts by two special lines cut parallel, making routing easier. The algorithm completely routes Burstein's switch-box problem and with an extension also routes the Deutsch channel example in 19 tracks     

TEE EFFECT OF EXTENDED THIN FILM EVAPORATION DURING TEE CONSTANT DRYING RATE PERIOD

The extended thin film evaporation is analyzed during the constant drying rate period. The extended thin film is defined as the liquid film which the disjoining pressure dominates the fluid flow field and works as the driving force of replenishing the evaporating liquid. The results of the analysis show that the extended thin film evaporation can compensate the reduction of evaporation rate due to the increase of dry spots and keep the evaporation rate nearly constant. Experimental data in the literature are in good agreement with the results of this work. This implies that the extended thin film evaporation may be an important part of the mechanism for drying.