Bi-criteria risk analysis of domain-specific and cross-domain changes in complex systems

2014 ◽  
Vol 73 ◽  
pp. 51-60 ◽  
Author(s):  
Kenneth H. Doerr ◽  
Keebom Kang
Keyword(s):  
Risk Analysis ◽  
Complex Systems ◽  
Domain Specific ◽  
Cross Domain

scholarly journals DISSECT: DISentangle SharablE ConTent for Multimodal Integration and Crosswise-mapping

2020 ◽  
Author(s):  
Geoffrey Schau ◽  
Erik Burlingame ◽  
Young Hwan Chang
Keyword(s):  
Deep Learning ◽  
Complete Information ◽  
Specific Information ◽  
Multimodal Integration ◽  
Specific Data ◽  
Domain Specific ◽  
Cross Domain ◽  
Input Feature ◽  
Novel Approach ◽  
Latent Representations

AbstractDeep learning systems have emerged as powerful mechanisms for learning domain translation models. However, in many cases, complete information in one domain is assumed to be necessary for sufficient cross-domain prediction. In this work, we motivate a formal justification for domain-specific information separation in a simple linear case and illustrate that a self-supervised approach enables domain translation between data domains while filtering out domain-specific data features. We introduce a novel approach to identify domainspecific information from sets of unpaired measurements in complementary data domains by considering a deep learning cross-domain autoencoder architecture designed to learn shared latent representations of data while enabling domain translation. We introduce an orthogonal gate block designed to enforce orthogonality of input feature sets by explicitly removing non-sharable information specific to each domain and illustrate separability of domain-specific information on a toy dataset.

A Transformer-Based Approach to Multilingual Fake News Detection in Low-Resource Languages

2022 ◽  
Vol 21 (1) ◽  
pp. 1-20
Author(s):  
Arkadipta De ◽  
Dibyanayan Bandyopadhyay ◽  
Baban Gain ◽  
Asif Ekbal
Keyword(s):  
Language Processing ◽  
English Language ◽  
Neural Model ◽  
Fake News ◽  
Web Content ◽  
Low Resource ◽  
Domain Specific ◽  
Cross Domain ◽  
Current State ◽  
Domain Transfer

Fake news classification is one of the most interesting problems that has attracted huge attention to the researchers of artificial intelligence, natural language processing, and machine learning (ML). Most of the current works on fake news detection are in the English language, and hence this has limited its widespread usability, especially outside the English literate population. Although there has been a growth in multilingual web content, fake news classification in low-resource languages is still a challenge due to the non-availability of an annotated corpus and tools. This article proposes an effective neural model based on the multilingual Bidirectional Encoder Representations from Transformer (BERT) for domain-agnostic multilingual fake news classification. Large varieties of experiments, including language-specific and domain-specific settings, are conducted. The proposed model achieves high accuracy in domain-specific and domain-agnostic experiments, and it also outperforms the current state-of-the-art models. We perform experiments on zero-shot settings to assess the effectiveness of language-agnostic feature transfer across different languages, showing encouraging results. Cross-domain transfer experiments are also performed to assess language-independent feature transfer of the model. We also offer a multilingual multidomain fake news detection dataset of five languages and seven different domains that could be useful for the research and development in resource-scarce scenarios.

On Developing Hybrid Modeling Methods using Metamodeling Platforms

2015 ◽  
Vol 6 (1) ◽  
pp. 47-66
Author(s):  
Srdjan Zivkovic ◽  
Krzystof Miksa ◽  
Harald Kühn
Keyword(s):  
Complex Systems ◽  
Hybrid Modeling ◽  
Modeling Languages ◽  
Modeling Tools ◽  
Analysis And Design ◽  
Domain Specific ◽  
Modeling Methods ◽  
Model Based ◽  
Domain Specific Modeling ◽  
Specific Modeling

It has been acknowledged that model-based approaches and domain-specific modeling (DSM) languages, methods and tools are beneficial for the engineering of increasingly complex systems and software. Instead of general-purpose one-size-fits-all modeling languages, DSM methods facilitate model-based analysis and design of complex systems by providing modeling concepts tailored to the specific problem domain. Furthermore, hybrid DSM methods combine single DSM methods into integrated modeling methods, to allow for multi-perspective modeling. Metamodeling platforms provide flexible means for design and implementation of such hybrid modeling methods and appropriate domain-specific modeling tools. In this paper, we report on the conceptualization of a hybrid DSM method in the domain of network physical devices management, and its implementation based on the ADOxx metamodeling platform. The method introduces a hybrid modeling approach. A dedicated DSM language (DSML) is used to model the structure of physical devices and their configurations, whereas the formal language for knowledge representation OWL2 is used to specify configuration-related constraints. The outcome of the work is a hybrid, semantic technology-enabled DSM tool that allows for efficient and consistency-preserving model-based configuration of network equipment.

scholarly journals Domain Conditioned Adaptation Network

2020 ◽  
Vol 34 (07) ◽  
pp. 11386-11393 ◽  
Author(s):  
Shuang Li ◽  
Chi Liu ◽  
Qiuxia Lin ◽  
Binhui Xie ◽  
Zhengming Ding ◽  
...  
Keyword(s):  
Domain Adaptation ◽  
Feature Learning ◽  
Feature Representations ◽  
Domain Specific ◽  
Cross Domain ◽  
Learning Tasks ◽  
Domain Learning ◽  
High Level ◽  
High Level Feature ◽  
Target Data

Tremendous research efforts have been made to thrive deep domain adaptation (DA) by seeking domain-invariant features. Most existing deep DA models only focus on aligning feature representations of task-specific layers across domains while integrating a totally shared convolutional architecture for source and target. However, we argue that such strongly-shared convolutional layers might be harmful for domain-specific feature learning when source and target data distribution differs to a large extent. In this paper, we relax a shared-convnets assumption made by previous DA methods and propose a Domain Conditioned Adaptation Network (DCAN), which aims to excite distinct convolutional channels with a domain conditioned channel attention mechanism. As a result, the critical low-level domain-dependent knowledge could be explored appropriately. As far as we know, this is the first work to explore the domain-wise convolutional channel activation for deep DA networks. Moreover, to effectively align high-level feature distributions across two domains, we further deploy domain conditioned feature correction blocks after task-specific layers, which will explicitly correct the domain discrepancy. Extensive experiments on three cross-domain benchmarks demonstrate the proposed approach outperforms existing methods by a large margin, especially on very tough cross-domain learning tasks.

scholarly journals Distinction of Domain-Specific and Cross-Domain Linkage Types for Engineering Change Management

2019 ◽  
Vol 1 (1) ◽  
pp. 1125-1134
Author(s):  
Robert Wilms ◽  
David Inkermann ◽  
Vadym Finn Cemmasson ◽  
Michael Reik ◽  
Thomas Vietor
Keyword(s):  
Change Management ◽  
Organizational Structures ◽  
Change Propagation ◽  
Engineering Change ◽  
Domain Specific ◽  
Engineering Change Management ◽  
Cross Domain ◽  
Product Models ◽  
Additional Costs ◽  
Engineering Changes

AbstractEngineering Changes (ECs) are substantial elements of the design process of technical products and are in particular relevant for companies due to enormous additional costs and time delays they can cause. In order to better understand ECs and realize efficient Engineering Change Management (ECM), different approaches exist. One aspect of ECM are change propagation analysis, which try to analyze knock-on effects of an EC on other product elements or the development process. How ECs can propagate is in particular difficult to assess for complex products realized within different engineering domains (mechanical, electrical and software engineering). To address this challenge, ECs are classified, strategies to cope with ECs are presented and change propagation approaches are analyzed in this paper. Thereby a lack of indicators for cross-domain propagation is identified. To overcome this issue, the distinction of domain-specific and cross-domain linkage types is proposed and a set of linkage types is presented. Further research is motivated to integrate these linkage types in product models while also considering processes and organizational structures as additional dimensions of ECM.

Outcomes of Organizational and Community Embeddedness: A Conservation of Resources Perspective

2020 ◽  
pp. 105960112096356
Author(s):  
Barjinder Singh ◽  
Margaret Shaffer ◽  
Thirumalai Thattai Rajan Selvarajan
Keyword(s):  
Underlying Mechanism ◽  
Future Research ◽  
Conservation Of Resources ◽  
Mediating Role ◽  
Domain Specific ◽  
Community Embeddedness ◽  
Cross Domain ◽  
Community Outcomes ◽  
Psychological Flourishing

Drawing on Conservation of Resources and spillover theories, we empirically examine work and community outcomes of both organizational and community embeddedness and the underlying mechanism whereby the two forms of embeddedness influence both domain-specific and cross-domain outcomes. With data from 165 matched pairs of employees and their colleagues from a Midwestern US organization, we found that organizational and community embeddedness influence specific individual behaviors both within and across their respective domains. Additionally, we found support for the mediating role of psychological flourishing in the relationships between embeddedness and various organizational and community outcomes. We discuss the theoretical contributions and practical implications of our findings, as well as suggestions for future research.

scholarly journals Can being scared make your tummy ache? Naive theories, ambiguous evidence and preschoolers’ causal inferences

2018 ◽  
Author(s):  
Laura Schulz ◽  
Elizabeth Bonawitz ◽  
Tom Griffiths
Keyword(s):  
Statistical Learning ◽  
Biological Effect ◽  
Bayesian Model ◽  
Statistical Evidence ◽  
Causal Learning ◽  
Causal Inferences ◽  
Domain Specific ◽  
Naive Theories ◽  
Cross Domain ◽  
General Statistical

Causal learning requires integrating constraints provided by domain-specific theorieswith domain-general statistical learning. In order to investigate the interaction between these factors, preschoolers were presented with stories pitting their existing theories against statistical evidence. Each child heard two stories in which two candidate causes co-occurred with an effect. Evidence was presented in the form: ABàE, ACàE, AD àE, etc. In one story, all variables came from the same domain; in the other, the recurring candidate cause, A, came from a different domain (A was a psychological cause of a biological effect). After receiving this statistical evidence, children were asked to identify the cause of the effect on a new trial. Consistent with the predictions of a Bayesian model, all children were more likely to identify A as the cause within domains than across domains. While three-and-half-year-olds learned only from the within- domain evidence, four- and five-year-olds learned from the cross-domain evidence and were able to transfer their new expectations about psychosomatic causality to a novel task.

scholarly journals Neural Networks Incorporating Unlabeled and Partially-labeled Data for Cross-domain Chinese Word Segmentation

2018 ◽  
Author(s):  
Lujun Zhao ◽  
Qi Zhang ◽  
Peng Wang ◽  
Xiaoyu Liu
Keyword(s):  
Large Scale ◽  
Word Segmentation ◽  
Language Models ◽  
Cross Entropy ◽  
Chinese Word ◽  
Chinese Word Segmentation ◽  
Domain Specific ◽  
Partially Labeled Data ◽  
Cross Domain ◽  
Resource Poor

Most existing Chinese word segmentation (CWS) methods are usually supervised. Hence, large-scale annotated domain-specific datasets are needed for training. In this paper, we seek to address the problem of CWS for the resource-poor domains that lack annotated data. A novel neural network model is proposed to incorporate unlabeled and partially-labeled data. To make use of unlabeled data, we combine a bidirectional LSTM segmentation model with two character-level language models using a gate mechanism. These language models can capture co-occurrence information. To make use of partially-labeled data, we modify the original cross entropy loss function of RNN. Experimental results demonstrate that the method performs well on CWS tasks in a series of domains.

Sign in / Sign up

Export Citation Format

Share Document