Knowledge Transfer Strategies Research Articles

Danza y universidad en tiempos de la Covid-19

The article presents the results of documentary research that reflects in a systematic way about virtual formative experiences, open to the public and led by dance programs at the university studies education in times of Covid-19. This global event that gained strength in 2020 impacted the training practices in all educational institutions and destabilized their knowledge transfer mechanisms. Therefore, the research is posed as questions: ?What are the knowledge transfer strategies that the Universities with Dance programs carried out during the year 2020? Likewise, what are the complementary contents and topics to the regular study plan that were privileged during said transfers? The conceptual basis on which the research is based involves the formative processes in university studies education and the dancing body, taking into consideration the statements of the World Health Organization, the Inter-American Development Bank, and the guidelines in university studies education raised by the Ministry of Education of Colombia.

A knowledge-guide hierarchical learning method for long-tailed image classification

Deep visual recognition methods have achieved excellent performance on artificially constructed image datasets where the data distribution is balanced. However, in real-world scenarios, data distribution is usually extremely imbalanced and exhibit a long-tailed distribution where data in each head class is more than the class in the tail. Many efficient deep learning methods fail to work normally, i.e., they perform well in the head class while poor in the tail class. In this paper, we propose a two-layer Hierarchical-Learning Long-Tailed Recognition (HL-LTR) algorithm which transforms the long-tailed problem into a hierarchical classification problem by constructing a hierarchical superclass tree in which each layer corresponds to a recognition task. In the first layer of the tree, the degree of data imbalance is largely decreased. The recognition task of the second layer is the original long-tailed recognition problem. The training of HL-LTR is top-down. The knowledge learned by the first layer transfers to classes of the second layer and guides the feature learning of the second layer by using attention mechanism module and knowledge distillation method. Compared with directly solving the most difficult long-tailed recognition task, HL-LTR achieves better performance due to its progressive learning method from easy to difficult and effective knowledge transfer strategy.

Enhancing business community disaster resilience. A structured literature review of the role of dynamic capabilities

PurposeThe purpose of the current literature reviews to contribute to the study of organisational resilience. The study intends to understand the role of dynamic capability flow in creating more resilient business communities. The study tries to relate to how continuous learning enables business communities to plan for, respond to and bounce back from disasters.Design/methodology/approachA structured literature review of 38 empirical studies indicates organisational resilience and highlights the dynamic attributes of organisational resilience and the importance of knowledge interactions and information sharing.FindingsContinuous learning results in complex dynamic capability manifested through research and innovations, technological implementations, social learnings and community knowledge sharing. This clearly emphasises the role of dynamic capabilities in fostering disaster resilience in organisations and business communities.Research limitations/implicationsResearch limitations on continuous learning can be cited as to incorporate case study methods related to organisational experiences during the COVID-19 pandemic while also reviewing digital knowledge transfer strategies and influencing factors.Practical implicationsThe study for fostering resilience in the business ecosystem needs to be embedded in the continuous learning process, and it also includes knowledge sharing and collaboration, both externally and internally, for the business community.Social implicationsSocial implications for this study relate to the seamless flow of the knowledge transfer and knowledge sharing process. The dynamic process of organisational and the business community resilience is a key outcome of this knowledge sharing and knowledge transfer process.Originality/valueA conceptual framework was developed from the review, emphasising how dynamic capabilities through continuous learning enhance business community resilience.

Exploring reverse knowledge transfer and asset augmentation strategy by developed country MNEs: Case study evidence from the Indian pharmaceutical industry

Using an in-depth qualitative longitudinal case study approach, we examine strategic asset and knowledge augmentation strategies of an advanced economy multinational enterprise (MNE). Our study is unique as it is contextualized in the knowledge-intensive pharmaceutical industry within the changing institutional landscape in India. And, in contrast to previous studies, it focuses on RKT from a newly acquired subsidiary where the protocol for knowledge transfer and relationship between MNE headquarters (HQ) and subsidiary does not exist. We contribute to the evolving literature on reverse knowledge transfer (RKT) which assumes that MNEs extract knowledge from their existing subsidiaries to strengthen their competitive advantages. We also reveal that MNEs design an architecture, that is reflected in the mechanism, governance structure, and timing, for efficient transfer and effective absorption of the knowledge and assets acquired. We introduce a novel concept of ‘reverse asset augmentation’ (RAA) to capture the MNE’s behavior of strategic asset seeking from emerging economies. RAA is parallel to and entwined with RKT conceptualization as knowledge is embedded within certain physical assets, other assets, such as brand name/image, may not embody knowledge. Thematic and processual analysis of interview data, collected in three phases, suggests that RAA complements RKT and that both concepts together capture the unconventional strategies of advanced economy MNEs acquiring emerging economy MNEs in search of strategic assets and knowledge.

A Neural Mechanism for Transfer Learning

Transfer learning, the application of knowledge from a training domain to a novel domain, is a critical advantage for human intelligence and an eagerly sought goal for artificial intelligence. Here, we describe a neural mechanism for transfer learning based on compositional representation. In monkeys trained on a small alphabet of letter-like stimuli, neurons in high-level visual cortex showed increased selectivity specifically for compositional letter elements (limbs, junctions, terminations) and their interactions (combinations of limbs, etc.) and not for whole letters. Strikingly, the enhanced selectivity for the familiar compositional elements and their interactions transferred to novel letters constructed from the same elements, making the novel letters more discriminable and providing information about their structure. Compositionality is a powerful strategy for knowledge transfer in domains like vision where objects often share common elements. Our results reveal a brain mechanism for visual transfer learning that could be adapted for artificial vision.

Variation in prophylactic tranexamic acid administration among anesthesiologists and surgeons in orthopedic surgery: a retrospective cohort study.

Tranexamic acid (TXA) reduces red blood cell transfusion in various orthopedic surgeries, yet the degree of practice variation in its use among anesthesiologists and surgeons has not been described. To target future knowledge transfer and implementation strategies, and to better understand determinants of variability in prophylactic TXA use, our primary objective was to evaluate the influence of surgical team members on the variability of prophylactic TXA administration. This was a retrospective cohort study of all adult patients undergoing primary total hip arthroplasty (THA), hip fracture surgery, and spine fusion ± vertebrectomy at two Canadian hospitals between January 2014 and December 2016. We used Canadian Classification of Health Interventions procedure codes within the Discharge Abstract Database which we linked to the Ottawa Data Warehouse. We described the percentage of patients that received TXA by individual surgery, the specifics of TXA dosing, and estimated the effect of anesthesiologists and surgeons on prophylactic TXA using multivariable mixed-effects logistic regression analyses. In the 3,900 patients studied, TXA was most commonly used in primary THA (85%; n = 1,344/1,582), with lower use in hip fracture (23%; n = 342/1,506) and spine fusion surgery (23%; n = 186/812). The median [interquartile range] total TXA dose was 1,000 [1,000-1,000] mg, given as a bolus in 92% of cases. Anesthesiologists and surgeons added significant variability to the odds of receiving TXA in hip fracture surgery and spine fusion, but not primary THA. Most of the variability in TXA use was attributed to patient and other factors. We confirmed the routine use of TXA in primary THA, while observing lower utilization with more variability in hip fracture and spine fusion surgery. Further study is warranted to understand variations in use and the barriers to TXA implementation in a broader population of orthopedic surgical patients at high risk for transfusion.

Knowledge Distillation with Contrastive Inter-Class Relationship

Due to the high computational cost, the application of deep neural networks (DNNs) to the real-time tasks has been limited. A possible solution is to compress the size of the model so that the demand for computation resources can be decreased. A popular method is called knowledge distillation (KD). The basic philosophy behind KD is to transfer the information extracted from the larger teacher network to the smaller student network. The general knowledge transfer strategy is to match the one-to-one logit or intermedia layers between the teacher and student networks, correspondingly. This objective may neglect the informative relationship information between different samples. In this paper, we borrow the idea of metric learning to transfer the contrastive relationship information learned from teacher network to the student. Specifically, we use the well-known Triplet loss to regularize the training of the student network. By the modified negative selection strategy, our Contrastive Knowledge Distillation (CKD) method can efficiently improve the performance of the student network compared with the traditional KD methods. Empirical experiments on KD benchmarks and real-world datasets also demonstrate the superiority of CKD.

Gesture Recognition in Robotic Surgery: A Review.

Surgical activity recognition is a fundamental step in computer-assisted interventions. This paper reviews the state-of-the-art in methods for automatic recognition of fine-grained gestures in robotic surgery focusing on recent data-driven approaches and outlines the open questions and future research directions. An article search was performed on 5 bibliographic databases with the following search terms: robotic, robot-assisted, JIGSAWS, surgery, surgical, gesture, fine-grained, surgeme, action, trajectory, segmentation, recognition, parsing. Selected articles were classified based on the level of supervision required for training and divided into different groups representing major frameworks for time series analysis and data modelling. A total of 52 articles were reviewed. The research field is showing rapid expansion, with the majority of articles published in the last 4 years. Deep-learning-based temporal models with discriminative feature extraction and multi-modal data integration have demonstrated promising results on small surgical datasets. Currently, unsupervised methods perform significantly less well than the supervised approaches. The development of large and diverse open-source datasets of annotated demonstrations is essential for development and validation of robust solutions for surgical gesture recognition. While new strategies for discriminative feature extraction and knowledge transfer, or unsupervised and semi-supervised approaches, can mitigate the need for data and labels, they have not yet been demonstrated to achieve comparable performance. Important future research directions include detection and forecast of gesture-specific errors and anomalies. This paper is a comprehensive and structured analysis of surgical gesture recognition methods aiming to summarize the status of this rapidly evolving field.

Learning Slimming SAR Ship Object Detector Through Network Pruning and Knowledge Distillation

The deployment of deep convolutional neural networks (CNNs) in synthetic aperture radar (SAR) ship real-time detection is largely hindered by huge computational cost. In this article, we propose a novel learning scheme for training a lightweight ship detector called Tiny YOLO-Lite, which simultaneously 1) reduces the model storage size; 2) decreases the floating point operations (FLOPs) calculation; and 3) guarantees the high accuracy with faster speed. This is achieved by self-designed backbone structure and network pruning, which enforces channel-level sparsity in the backbone network and yields a compact model. In addition, knowledge distillation is also applied to make up for the performance decline caused by network pruning. Hereinto, we propose to let small student model mimic cumbersome teacher's output to achieve improved generalization. Rather than applying vanilla full feature imitation, we redefine the distilled knowledge as the inter-relationship between different levels of feature maps and then transfer it from the large network to a smaller one. On account that the detectors should focus more on the salient regions containing ships while background interference is overwhelming, a novel attention mechanism is designed and then attached to the distilled feature for enhanced representation. Finally, extensive experiments are conducted on SSDD, HRSID, and two large-scene SAR images to verify the effectiveness of the thinner SAR ship object detector in comparison of with other CNN-based algorithms. The detection results demonstrate that the proposed detector can achieve lighter architecture with 2.8-M model size, more efficient inference (>200 fps) with low computation cost, and more accurate prediction with knowledge transfer strategy.

Increasing the impact of science and technology to provide more people with healthier and safer food

AbstractEnsuring adequate food availability to an increasing world population constitutes one of the biggest challenges faced by humankind. Scientific and technological advances in food production during the last century enabled agriculture to cope with the concomitant increase in food demand. For example, cereal yields have more than doubled from a global average of 1.5 metric tons per hectare in the 1960s up to 3.2 metric tons per hectare in 2018. This was made possible by the work in different research fields such as agronomy, engineering, and plant sciences, showing that an inter and multidisciplinary approach is indispensable for significant progress. This manuscript is aimed at generating reflexion and analysis about the challenges that agriculture faces at present to satisfy projected food demands, which implies a further doubling of food production by 2050, according to the latest estimates. Relevant issues related to food production (climate change, pollution of water and soils by pesticides and fertilizers, loss of germplasm and biodiversity) are discussed and potential solutions to achieve food security in quantity and quality are reviewed, mainly from the plant breeding and crop‐production perspectives, always associated with environmental health preservation and improvement. A broad transdisciplinary effort is needed to increase the impact of science and technology to provide more people with healthier and safer food, produced in a sustainable way. Nonetheless, science and technology alone will not succeed to meet those challenges. Education and knowledge transfer strategies are needed to guarantee responsible production and consumption everywhere, therefore allowing the benefits of scientific and technological progress reach the world population. Simultaneously, adequate action by regulatory authorities and governments concerted at international level, with thorough application of the Precautionary Principle, and aiming at environmental and social justice are imperatively required to meet the challenge and achieve the goal.

Regularized Evolutionary Multitask Optimization: Learning to Intertask Transfer in Aligned Subspace

This article proposes a novel and computationally efficient explicit intertask information transfer strategy between optimization tasks by aligning the subspaces. In evolutionary multitasking, the tasks might have biases embedded in function landscapes and decision spaces, which often causes the threat of predominantly negative transfer. However, the complementary information among different tasks can give an enhanced performance of solving complicated problems when properly harnessed. In this article, we distill this insight by introducing an intertask knowledge transfer strategy implemented in the low-dimension subspaces via a learnable alignment matrix. Specifically, to unveil the significant features of the function landscapes, the task-specific low-dimension subspaces is established based on the distribution information of subpopulations possessed by tasks, respectively. Next, the alignment matrix between pairwise subspaces is learned by minimizing the discrepancies of the subspaces. Given the aligned subspaces by applying the alignment matrix to subspaces' base vectors, the individuals from different tasks are then projected into aligned subspaces and reproduce therein. Moreover, since this method only considers the leading eigenvectors, it turns out to be intrinsically regularized and noise-insensitive. Comprehensive experiments are conducted on the synthetic and practical benchmark problems so as to assess the efficacy of the proposed method. According to the experimental results, the proposed method exhibits a superior performance compared with existing evolutionary multitask optimization algorithms.

Semantic strategies in ubiquitous music: Deploying the sound sphere ecology in transitional settings

We report the results of a study involving twenty subjects doing musical activities in transitional settings, supported by an ecology of tools based on the metaphor for creative action Sound Sphere. The Sound Sphere Ecology (SFS) is a set of web-based tools, loosely organized around audio mixing and processing tasks. It employs verbal strategies for knowledge transfer to provide support for lay participants and specialists. To understand how the stakeholders influence and are influenced by this design strategy, we carried out a series of experiments involving assessments of the participants' behaviours and of the sonic products during various creative musical tasks with SFS. The overall results were positive, indicating that the proposed metaphor provides effective support for casual interaction, highlighting the participants' level of engagement. As a downside, the assessments pointed to ease of use as the lowest and less consistent item among the rated creative factors. We discuss the implications of these results and propose various design enhancements to enable the usage of a larger pool of resources. Considering the heterogeneous profiles of casual stakeholders, methodological refinements are also proposed to assess the knowledge gained by the participants during the exploratory activities, while augmenting their ability to share knowledge. This is one of the first studies on creativity-action metaphors for casual interaction.

Deep Domain Generalization Combining A Priori Diagnosis Knowledge Toward Cross-Domain Fault Diagnosis of Rolling Bearing

Recent works suggest that using knowledge transfer strategies to tackle cross-domain diagnosis problems is promising for achieving engineering diagnosis. This article presents a diagnosis scheme for rolling bearing under a challenging domain generalization scenario, in which more potential discrepancies among multiple source domains are eliminated and only normal samples of the target domain are available during the training stage. To achieve sufficient generalization performance, a diagnosis scheme combining some a priori diagnosis knowledge and a deep domain generalization network for fault diagnosis (DDGFD) is elaborated. Through signal preprocessing steps guided by the a priori diagnosis knowledge, the inputs of DDGFD with a primary consistent meaning across domains are constructed from the vibration signal. On this basis, DDGFD would intently release its talent on learning discriminative and domain-invariant fault features from source domains, and then generalize the learned knowledge to identify unseen target samples. On cross-domain tasks organized using broad bearing data sets, the superiority of DDGFD is validated by comparing its performance with various data-driven diagnosis methods.

A novel approach to learning virtual engineering in the medical devices process

Currently, the use of computer-based technologies has increased to support the design process of biomedical products. However, the limited capacity of undergraduates to work with the technologies used in the development process negatively influences the research execution for biomedical product design. For this reason, we design a framework for knowledge transfer required by undergraduate students. The study aim was to evaluate the framework proposed as a strategy for knowledge transfer and ICT skills strengthening. The framework presented, manage teaching and learning process of software tools used during the ideation and development of biomedical products. The framework is oriented to undergraduates’ students in the research group, working in their final degree project. An evaluation tool was designed and applied with four undergraduate cohorts. The students’ perception of learning process satisfaction, complexity, difficulty, and usefulness of software applications was evaluated. The software applications evaluated were used in Reverse Engineering, Computer-Aided Design, Computer-Aided Engineering, and Additive Manufacturing. The results show: the role of stakeholders involved in the training process, perceived differences of teaching and learning strategy, and the value of technical support. Likewise, the empowerment and the affinity degree in developing biomedical products was evidenced.

Artificial Chemist: An Autonomous Quantum Dot Synthesis Bot.

The optimal synthesis of advanced nanomaterials with numerous reaction parameters, stages, and routes, poses one of the most complex challenges of modern colloidal science, and current strategies often fail to meet the demands of these combinatorially large systems. In response, an Artificial Chemist is presented: the integration of machine-learning-based experiment selection and high-efficiency autonomous flow chemistry. With the self-driving Artificial Chemist, made-to-measure inorganic perovskite quantum dots (QDs) in flow are autonomously synthesized, and their quantum yield and composition polydispersity at target bandgaps, spanning 1.9 to 2.9 eV, are simultaneously tuned. Utilizing the Artificial Chemist, eleven precision-tailored QD synthesis compositions are obtained without any prior knowledge, within 30 h, using less than 210 mL of total starting QD solutions, and without user selection of experiments. Using the knowledge generated from these studies, the Artificial Chemist is pre-trained to use a new batch of precursors and further accelerate the synthetic path discovery of QD compositions, by at least twofold. The knowledge-transfer strategy further enhances the optoelectronic properties of the in-flow synthesized QDs (within the same resources as the no-prior-knowledge experiments) and mitigates the issues of batch-to-batch precursor variability, resulting in QDs averaging within 1 meV from their target peak emission energy.

Assessing care models implemented in primary healthcare for persons with dementia: a mixed-methods study protocol

IntroductionDementia is on the rise in Canada and globally. Ensuring accessibility to diagnosis, treatment and management throughout the course of the disease is a very significant problem worldwide. In order...

Knowledge transfer strategies and practices for higher education institutions

PurposeThis study aims to identify effective strategies and practices for higher education institutions (HEIs) to enhance their knowledge recipients’ knowledge absorptive capacity.Design/methodology/approachQuestionnaire surveys and interviews were conducted to examine the effectiveness of a knowledge transfer (KT) project and these were administered by HEIs in Hong Kong. Pre- and post-test questionnaire surveys with t-tests were used to evaluate changes in the knowledge absorptive capacity of 1,014 participants from 20 schools. Qualitative interviews were conducted to investigate school leaders’ perception of the effectiveness of the KT strategies.FindingsResults revealed significant differences between the pre- and post-tests in all the elements of absorptive capacity. The processes of knowledge acquisition, contextualisation, internalisation and externalisation in a closed-loop mechanism were identified as effective KT strategies. Conducting training programmes, workshops, consultations, work-based studies and presentation seminars were found to be effective KT practices to support the recipients in acquiring, contextualising, internalising and externalising knowledge.Practical implicationsHEIs should provide consultative support to recipients by conducting work-based studies and presentations to enhance their knowledge absorptive capacities.Originality/valueThis study contributes empirical evidence to validate the application of Liyanage et al.’s (2009) KT model to HEIs’ KT model for the purpose of designing KT activities and enhancing the absorption capacities of the recipients. This research contributes an empirical closed-loop KT model, effective KT strategies and practices for HEIs to support their knowledge recipients so that they can internalise the acquired knowledge.

A Preliminary Study of Knowledge Transfer in Multi-Classification Using Gene Expression Programming.

Gene Expression Programming (GEP), a variant of Genetic Programming (GP), is a well established technique for automatic generation of computer programs. Due to the flexible representation, GEP has long been concerned as a classification algorithm for various applications. Whereas, GEP cannot be extended to multi-classification directly, and thus is only capable of treating an M-classification task as M separate binary classifications without considering the inter-relationship among classes. Consequently, GEP-based multi-classifier may suffer from output conflict of various class labels, and the underlying conflict can probably lead to the degraded performance in multi-classification. This paper employs evolutionary multitasking optimization paradigm in an existing GEP-based multi-classification framework, so as to alleviate the output conflict of each separate binary GEP classifier. Therefore, several knowledge transfer strategies are implemented to enable the interation among the population of each separate binary task. Experimental results on 10 high-dimensional datasets indicate that knowledge transfer among separate binary classifiers can enhance multi-classification performance within the same computational budget.

Codifying explicit and tacit practitioner knowledge in community social pediatrics organizations: Evaluation of the first step of a knowledge transfer strategy

In this article, we the present the evaluation of the first phase of a KT initiative, the codification of the tacit and explicit knowledge of community social pediatrics (CSP) practitioners. As increasing numbers of professionals began working in the CSP field, it became necessary to capture the know-how to make it better understood. This first phase of the initiative took advantage of one of the KM stages, the codifying process that is absent from KT models. The evaluation aim to answer the question: “What is the CSPs practitioners level of familiarity with the codification document, as well as their perceptions of its utility and its contribution to the understanding of the CSP model”. The results, show that most respondents had consulted the codification document and were able to describe in detail parts of it. The results also show that they considered it to be a useful and necessary tool to guide practices and helpful in understanding the model. The results of this evaluation is paving the way for the next steps of the KT initiative.

Syntactically Meaningful and Transferable Recursive Neural Networks for Aspect and Opinion Extraction

In fine-grained opinion mining, extracting aspect terms (a.k.a. opinion targets) and opinion terms (a.k.a. opinion expressions) from user-generated texts is the most fundamental task in order to generate structured opinion summarization. Existing studies have shown that the syntactic relations between aspect and opinion words play an important role for aspect and opinion terms extraction. However, most of the works either relied on predefined rules or separated relation mining with feature learning. Moreover, these works only focused on single-domain extraction, which failed to adapt well to other domains of interest where only unlabeled data are available. In real-world scenarios, annotated resources are extremely scarce for many domains, motivating knowledge transfer strategies from labeled source domain(s) to any unlabeled target domain. We observe that syntactic relations among target words to be extracted are not only crucial for single-domain extraction, but also serve as invariant “pivot” information to bridge the gap between different domains. In this article, we explore the constructions of recursive neural networks based on the dependency tree of each sentence for associating syntactic structure with feature learning. Furthermore, we construct transferable recursive neural networks to automatically learn the domain-invariant fine-grained interactions among aspect words and opinion words. The transferability is built on an auxiliary task and a conditional domain adversarial network to reduce domain distribution difference in the hidden spaces effectively in word level through syntactic relations. Specifically, the auxiliary task builds structural correspondences across domains by predicting the dependency relation for each path of the dependency tree in the recursive neural network. The conditional domain adversarial network helps to learn domain-invariant hidden representation for each word conditioned on the syntactic structure. In the end, we integrate the recursive neural network with a sequence labeling classifier on top that models contextual influence in the final predictions. Extensive experiments and analysis are conducted to demonstrate the effectiveness of the proposed model and each component on three benchmark data sets.