Find Research Output

Research Output
  • All
  • Scholar Profiles
  • Research Units
  • Research Output
Filter
Department Publication Year Content Type Data Sources

SELECTED FILTERS

2020
CPCI-SSH
Clear all

1.Corpora for document-level neural machine translation

Author:Liu,Siyou;Zhang,Xiaojun

Source:LREC 2020 - 12th International Conference on Language Resources and Evaluation, Conference Proceedings,2020,Vol.

Abstract:Instead of translating sentences in isolation, document-level machine translation aims to capture discourse dependencies across sentences by considering a document as a whole. In recent years, there have been more interests in modelling larger context for the state-of-the-art neural machine translation (NMT). Although various document-level NMT models have shown significant improvements, there nonetheless exist three main problems: 1) compared with sentence-level translation tasks, the data for training robust document-level models are relatively low-resourced; 2) experiments in previous work are conducted on their own datasets which vary in size, domain and language; 3) proposed approaches are implemented on distinct NMT architectures such as recurrent neural networks (RNNs) and self-attention networks (SANs). In this paper, we aim to alleviate the low-resource and under-universality problems for document-level NMT. First, we collect a large number of existing document-level corpora, which covers 7 language pairs and 6 domains. In order to address resource sparsity, we construct a novel document parallel corpus in Chinese-Portuguese, which is a non-English-centred and low-resourced language pair. Besides, we implement and evaluate the commonly-cited document-level method on top of the advanced Transformer model with universal settings. Finally, we not only demonstrate the effectiveness and universality of document-level NMT, but also release the preprocessed data, source code and trained models for comparison and reproducibility.

2.Application of Robotic Technologies for the Fabrication Of Traditional Chinese Timber Joints

Author:Zhao, JY;Lombardi, D;Agkathidis, A

Source:ECAADE 2020: ANTHROPOLOGIC - ARCHITECTURE AND FABRICATION IN THE COGNITIVE AGE, VOL 2,2020,Vol.

Abstract:The traditional Chinese building design was influenced by the climate and the sociogeographical conditions of the different regions in China. They were usually constructed out of wood relying on timber-joint based construction systems. Amongst the wide variety of the structural elements, the Dougong (bucket arch) is one of the most common components of traditional wooden framework buildings, presenting a high level of complexity. Parametric design and robotic technology enable new possibilities regarding its fabrication and application in contemporary architecture. Our paper will explore how the Dougong components could be reinvented through the use of parametric tools and robotic fabrication methods and thus applied to contemporary architectural structures. We will analyse and compare the properties of the original Dougong with the reinvented unit by using finite element analysis and digital optimization tools. Our findings will provide an insight into the traditional construction principles of the joint and how these can inform a design and fabrication framework for its application in contemporary buildings.

3.Smart Contracts for Decentralised Building Information Modelling

Author:Dounas, T;Jabi, W;Lombardi, D

Source:ECAADE 2020: ANTHROPOLOGIC - ARCHITECTURE AND FABRICATION IN THE COGNITIVE AGE, VOL 2,2020,Vol.

Abstract:The paper presents a model for decentralizing building information modelling, through implementing its infrastructure using the decentralized web. We discuss the shortcomings of BIM in terms of its infrastructure, with a focus on tracing identities of design authorship in this collective design tool. In parallel we examine the issues with BIM in the cloud and propose a decentralized infrastructure based on the Ethereum blockchain and the Interplanetary filesystem (IPFS). A series of computing nodes, that act as nodes on the Ethereum Blockchain, host disk storage with which they participate in a larger storage pool on the Interplanetary Filesystem. This storage is made available through an API is used by architects and designers creating and editing a building information model that resides on the IPFS decentralised storage. Through this infrastructure central servers are eliminated, and BIM libraries and models can be shared with others in an immutable and transparent manner. As such Architecture practices are able to exploit their intellectual property in novel ways, by making it public on the internet. The infrastructure also allows the decentralised creation of a resilient global pool of data that allows the participation of computation agents in the creation and simulation of BIM models.

4.Reflective Practice: Lessons Learnt by Using Board Games as a Design Tool for Location-Based Games

Author:Jones, C;Papangelis, K

Source:GEOSPATIAL TECHNOLOGIES FOR LOCAL AND REGIONAL DEVELOPMENT,2020,Vol.

Abstract:Location-based gaming (LBG) apps present many challenges to the design process. They have very different requirements compared to games that are aspatial in nature. They take place in the wild and this brings unique challenges to the practicalities of their design. There is a need to balance the core game play with the spatial requirements of location-aware technologies as well as considering the overall theme and objectives of the game together with the motivations and behaviours of players. We reflect upon this balancing act and explore an approach to creative paper prototyping through the medium of board games to co-design LBG requirements. We examine two case studies of location-based games with different goals. The first case study discusses the CrossCult Pilot 4 app built to trigger reflection on historical stories through thoughtful play. Whilst the second case study uses the City Conquerer app designed and played in Suzhou, China with a view to exploring notions of territoriality. The paper considers how spatial, social and interaction metaphors are used to simulate location-based games in a board game and discusses the lessons learned when transforming the paper game into a digital prototype. It forms part of a thinking by doing approach. By comparing the board games to the technical counterparts, we consider how effective are the features and activities implemented in the technology prototypes. We propose a set of 11 design constraints that developers must be mindful of when transitioning from paper to digital prototypes.

5.Multi-Objective Optimization for Daylight Retrofit

Author:De Luca, F;Wortmann, T

Source:ECAADE 2020: ANTHROPOLOGIC - ARCHITECTURE AND FABRICATION IN THE COGNITIVE AGE, VOL 1,2020,Vol.

Abstract:In sustainable building design, daylight improves occupants' wellbeing and reduces electric lighting use, but glazed areas can increase energy consumption for heating and cooling. Conflicting objectives such as daylight and energy consumption are the primary motivation behind multi-objective optimization. This paper presents the multi-objective optimization problem of maximizing daylight availability and minimizing whole energy consumption for the daylight retrofit of Tallinn University of Technology assembly hall, currently windowless. We present benchmark results of six different multi-objective algorithms and analyze the solutions on the best-known Pareto front. The majority of the analyzed solutions allow for adequate daylight provision of the building without additional energy consumption. Results of daylight and energy simulations for the analyzed solutions, are presented and discussed.

6.Aerodynamic Shape Optimization for High-Rise Conceptual Design Integrating and validating parametric design, (fast) fluid dynamics, structural analysis and optimization

Author:Zhang, R;Waibel, C;Wortmann, T

Source:ECAADE 2020: ANTHROPOLOGIC - ARCHITECTURE AND FABRICATION IN THE COGNITIVE AGE, VOL 1,2020,Vol.

Abstract:Using an integrated workflow with parametric design, Computational Fluid Dynamic (CFD) and Fast Fluid Dynamic (FFD) simulations, structural analysis and optimization, this paper evaluates the relative suitability of CFD and FFD simulations for Aerodynamic Shape Optimization (ASO). Specifically, it applies RBFOpt, a model-based optimization algorithm, to the ASO of a supertall high-rise. The paper evaluates the accuracy of the CFD and FDD simulations relative to a slower, more exact CFD simulation, and the performance of the model-based optimization algorithm relative to CMA-ES, an evolutionary algorithm. We conclude that FFD is useful for relative comparisons, such as for optimization, but less accurate than CFD in terms of absolute quantities. Although results tend to be similar, CMA-ES performs less well than RBFOpt for both large and small numbers of simulations, and for both CFD and FFD. RBFOpt with FFD emerges as the most suitable method for conceptual design, as it is much faster and only slightly less effective than RBFOpt with CFD.

7.Investor protection and the value impact of stock liquidity

Author:Huang, T;Wu, F;Yu, J;Zhang, BH

Source:JOURNAL OF INTERNATIONAL BUSINESS STUDIES,2020,Vol.51

Abstract:This paper investigates the effect of investor protection on the value impact of stock liquidity. Using a sample of firms from 40 countries for the period between 1996 and 2010, we show that investor protection is positively associated with the value impact of stock liquidity. This association is robust to the difference-in-differences approach based on a natural experiment. Further evidence shows that the positive effect of home-country investor protection on the liquidity-valuation association attenuates in countries with globally integrated capital markets.
Total 7 results found
Copyright 2006-2020 © Xi'an Jiaotong-Liverpool University 苏ICP备07016150号-1 京公网安备 11010102002019号