Exploiting Bayesian networks for fault isolation: A diagnostic case study of diesel fuel injection system

Fault isolation is known to be a challenging problem in machinery troubleshooting. It is not only because the isolation of multiple faults contains considerable number of uncertainties due to the strong correlation and coupling between different faults, but often massive prior knowledge is needed as well. This paper presents a Bayesian network-based approach for fault isolation in the presence of the uncertainties. Various faults and symptoms are parameterized using state variables, or the so-called nodes in Bayesian networks (BNs). Probabilistically causality between a fault and a symptom and its quantization are described respectively by a directed edge and conditional probability. To reduce the qualitative and quantitative knowledge needed, particular considerations are given to the simplification of Bayesian networks structures and conditional probability expressions using rough sets and noisy-OR/MAX model, respectively. By adopting the simplified approach, symptoms under multiple-fault are decoupled into the ones under every single fault, while the quantity of the conditional probabilities is simplified into the linear form of the faults quantity. Prior knowledge needed in Bayesian network-based diagnostic model is reduced significantly, which decreases the complexity in establishing and applying this diagnosis model. The computational efficiency is improved accordingly in the simplified BN model, after eliminating the redundant symptoms. The fault isolation methodology is illustrated through an example of diesel engine fuel injection system to verify the developed model.     

How does external knowledge sourcing enhance product development? Evidence from drug commercialization

Like in many industries, knowledge management is the key to the successful developments of new drugs and other commercialized products in the pharmaceutical sector. Prior studies have shown that knowledge management requires effective external knowledge sourcing strategies. We argue that both the geographic distance to external knowledge sources and the diversity of such sources have strong influences on the success of the search process and the development of new drugs. We test these theoretical assumptions by applying several advanced estimation techniques on a combined broad set of data from the U.S. Patent and Trademarks Office (USPTO) and U.S. Food and Drug Administration (FDA). We find empirical evidence that the intensity of external knowledge sourcing activities has a positive effect on new drugs approval and commercialization. Specifically, our models predict that a higher number of references to prior patents increases the probability of a new patented drug to be approved and commercialized. Furthermore, the geographic distance covered in the search for external knowledge also positively contributes to drug commercialization. Important management implications derived from these findings are also discussed in the paper.     

Topic-centric and semantic-aware retrieval system for internet of things

The Internet of things (IoT) has been considered as one of the promising paradigms that can allow people and objects to seamlessly interact. So far, numerous applications and services have been proposed, such as retrieval service. The retrieval, however, faces a big challenge in IoT because the data belongs to different domains and user interaction with the surrounding environment is constrained. This paper proposes Acrost, a retrieval system based on topic discovery and semantic awareness in IoT environment. The initial contents with interesting information is obtained through the combination of two topic centric collectors. The metadata is extracted by aggregating regular expression-based and conditional random fields-based approaches. Moreover, the semantic-aware retrieval is achieved by parsing the query and ranking the relevance of contents. In addition, we present a case study on academic conference retrieval to validate the proposed approaches. Experimental results show that the proposed system can significantly improve the response time and efficiency of topic self-adaptive retrieval manner.     

A meta-model for knowledge configuration management to support collaborative engineering

This paper focuses on industrial design and simulation processes especially in automotive and aerospace areas. Designers use business models (called expert models) such as CAD (computed aided design) and CAE (computed aided engineering) models to optimize and streamline the engineering process. Each expert model contains information such as parameters, expert rules, mathematic relations (parametric models, for example) which are shared by several users and in several different domains (mechanical, thermal, acoustic, fluid, etc.). This information is exploited at the same time in a concurrent engineering context. It is the basis of an imperfect collaboration process due to the fact that existing tools do not manage encapsulated information well and are unable to ensure that parameters and rules are consistent (same value of parameters for example) throughout different heterogeneous expert models. In this context, we propose an approach to manage knowledge using configurations synchronized with expert models which enable designers to use parameters consistently in a collaborative context. Our approach is called KCModel (knowledge configuration model): it allows acquisition, traceability, re-use and consistency of explicit knowledge used in configuration.     

Short- and long-term effects of students’ self-directed metacognitive prompts on navigation behavior and learning performance

This study seeks to promote learning in computer-based learning environments utilizing students’ self-directed metacognitive prompts. Such prompts are based on the idea of instructing students to design their own metacognitive scaffolds and learn with them afterward. In a pre-post experimental design, students in the experimental group (n = 35) were instructed to configure their own metacognitive prompts before learning whereas students in the control group (n = 35) learned without prompts. Log file analysis of navigation behavior indicates that students who learned with their individually designed, self-directed prompts visited relevant webpages significantly more often and spent a longer time on them compared with students in the control group. Moreover, participants in the experimental group attained better transfer performance immediately after learning. The long-term effect in transfer performance was even greater in a follow-up learning session conducted after three weeks without any instructional support in either group. These results are consistent with theories of metacognition and self-regulated learning and indicate that self-directed prompts can lead to sustainable effects.     

Automated generation of multiphysics simulation models to support multidisciplinary design optimization

To ensure a consistent design representation for serving multidisciplinary analysis, this research study proposes an intelligent modeling system to automatically generate multiphysics simulation models to support multidisciplinary design optimization processes by using a knowledge based engineering approach. A key element of this system is a multiphysics information model (MIM), which integrates the design and simulation knowledge from multiple engineering domains. The intelligent modeling system defines classes with attributes to represent various aspects of physical entities. Moreover, it uses functions to capture the non-physical information, such as control architecture, simulation test maneuvers and simulation procedures. The challenge of system coupling and the interactions among the disciplines are taken into account during the process of knowledge acquisition. Depending on the domain requirements, the intelligent modeling system extracts the required knowledge from the MIM and uses this first to instantiate submodels and second to construct the multiphysics simulation model by combining all submodels. The objective of this research is to reduce the time and effort for modeling complex systems and to provide a consistent and concurrent design environment to support multidisciplinary design optimization. The development of an unstable and unmanned aerial vehicle, a multirotor UAV, is selected as test case. The intelligent modeling system is demonstrated by modeling thirty-thousand multirotor UAV designs with different topologies and by ensuring the automatic development of a consistent control system dedicated for each individual design. Moreover, the resulting multiphysics simulation model of the multirotor UAV is validated by comparing with the flight data of an actual quadrotor UAV. The results show that the multiphysics simulation model matches test data well and indicate that high fidelity models can be generated with the automatic model generation process.     

Misinformed About the Affordable Care Act? Leveraging Certainty to Assess the Prevalence of Misperceptions

According to some recent research, Americans hold a great deal of misinformation about important political issues. However, such investigations treat incorrect answers to quiz questions measuring knowledge as evidence of misinformation. This study instead defines misperceptions as incorrect answers that respondents are confident are correct. Two surveys of representative samples of American adults on the Affordable Care Act reveal that most people were uncertain about the provisions in the law. Confidently held incorrect beliefs were far less common than incorrect answers. Misperceptions were most prevalent on aspects of the law on which elites prominently and persistently made incorrect claims. Furthermore, although Americans appear to have learned about the law between 2010 and 2012, misperceptions on many provisions of the law persisted.     

ERP system usage and benefit: A model of antecedents and outcomes

ERP system usage has been identified as a critical factor in attaining the benefit from an ERP installation. However, the specific antecedents of ERP usage and its impact on ERP benefit remains largely unknown. Drawing on absorption capacity theory, this study develops a theoretical model that examines the mediating effect of ERP system usage on ERP benefits. Similarly, the study also identifies the antecedents of ERP system usage. A model is tested using the responses of 157 ERP system end-users across the United States and the results suggest that ERP system usage is directly related ERP benefit. However, the relationship is moderated by the degree of knowledge integration mechanisms within the firm. Consistent with the proposed model, the results also reveal that technical resources, organizational fit and the extent of ERP implementation are key drivers of ERP system usage. The research findings advance our knowledge on how managers can enhance ERP usage and realize optimal ERP benefits.     

A hybrid approach based on fuzzy DEMATEL and FMCDM to predict success of knowledge management adoption in supply chain

Knowledge management (KM) adoption in the supply chain (SC) needs high investment as well as few changes in culture of entire SC. This study proposes a prediction framework based on the fuzzy decision-making trail and evaluation laboratory (DEMATEL) and fuzzy multi-criteria decision-making (FMCDM) for KM adoption in SC. This study first identifying the evaluation criteria of KM adoption in SC from literature review and expert opinion. Further, it uses fuzzy DEMATEL to evaluate weighting of each evaluation criteria's, after that FMCDM method uses to obtain possible rating of success of KM adoption in SC. The proposed approach is helpful to predict the success of KM adoption in SC without actually adopted KM in SC. It also enables organizations to decide whether to initiate KM, restrain adoption or undertake remedial improvements to increase the possibility of successful KM adoption in SC. This prominent advantage can be considered as one of the contribution of this paper. This proposed approach demonstrated with empirical case of a hydraulic valve manufacturing organization in India.     

基于知识图谱的用户偏好神经建模框架

针对目前基于特征和基于路径的知识图谱感知推荐方法的不足,文中提出端到端的将知识图谱引入推荐系统的用户偏好神经建模框架(NUPM).该框架以用户在知识图谱中的历史访问项目为偏好起点,通过知识图谱中实体间的关系链接传播用户偏好,学习用户的潜在偏好,同时使用注意力网络融合各传播阶段偏好特征以构建最终的用户偏好向量.在真实数据集上的对比实验表明文中框架在个性化推荐中对用户偏好刻画的有效性.