Joint Knowledge Research Articles

Empowering Knowledge Distillation via Open Set Recognition for Robust 3D Point Cloud Classification

• We propose a joint training methodology for knowledge distillation and open set recognition to improve model robustness. • We demonstrate how knowledge distillation (KD) and open set recognition (OSR) can be performed on 3D point cloud data • The proposed method provides a much smaller model which additionally has enhanced OSR capabilities than the original model. • We find that KD by itself can transfer OSR ability to a student model apart from previously known dark knowledge transfer. • We show how there is a trade-off that comes into play for the student model’s open-set and closed-set performance. Real-world scenarios pose several challenges to deep learning based computer vision techniques despite their tremendous success in research. Deeper models provide better performance, but are challenging to deploy and knowledge distillation allows us to train smaller models with minimal loss in performance. A model also has to deal with open set samples from classes outside the ones it was trained on and should be able to identify them as unknown samples while classifying the known ones correctly. Finally, most existing image recognition research focuses only on using two-dimensional snapshots of the three-dimensional real world objects. In this work, we attempt to bridge these three research fields, which have been developed independently until now, despite being deeply interrelated in practice. We propose a joint knowledge distillation and open set recognition training methodology for three-dimensional object recognition. We demonstrate the effectiveness of the proposed method via various experiments on how it allows us to obtain a much smaller model, which takes a minimal hit in performance while being capable of open set recognition for 3D point cloud data.

Transforming construction: heterarchical megaproject ecologies and the management of innovation

As the digital transformation impacts many sectors of the economy, actors in the construction industry—and more specifically in megaproject initiatives—have to adapt to new technologies and processes. Megaprojects are commonly undertaken to build essential infrastructures such as roads, dams, buildings, or even smart cities or districts, and usually involve complex and hybrid organizational forms. Moreover, digitalization transforms megaprojects, presenting megaproject teams with opportunities, but also challenges. This conceptual paper explores the characteristics of heterarchical megaproject ecologies in order to identify ways to address the impacts of the current digital transformation. Our aim is to analyze how a heterarchical form of governance contributes to transforming the management of innovation. While heterarchies can lead to severe dysfunctions, we propose strategies to manage them, paying specific attention to the governance of a common-pool-resource scenario, network roles, knowledge articulation and learning. The main contribution of this paper is to provide a renewed conceptualization of megaproject governance and to propose a conceptual framework that can be used to study the management of innovation in empirical megaproject settings.

University-Industry R&D Alliances – An Exploration of Alliance Capabilities Measures

Background: Alliance capabilities studies have long emerged since the 1990s, focusing mainly on firm-to-firm collaboration. However, research on university-industry alliances only emerged from the 2000s. Alliance capabilities are portrayed as a crucial condition to achieve the targeted collaboration outcomes and sustainable relationships. As most alliance capabilities studies focus on firm-to-firm collaboration, research on university-industry research and development (R&D) alliance is still scarce. Thus, the measurement items for alliance capabilities in the university-industry R&D context are still under-developed. Thus, to investigate how alliance capabilities affect university-industry R&D performance in Malaysia, the relevant measures must first be defined. This paper intends to properly define the measurement items for alliance capabilities in the context of university-industry R&D alliances. Methodology: The alliance capabilities measures are adapted from various literature to accommodate both university and industry perspectives. In finalizing the measurement, in-depth pre-testing was conducted by five strategic management subject matter experts in ensuring face and content validity. Results: There are three alliance capability dimensions. The first dimension is alliance management capability which includes goal setting, process configuration, alliance structure, coordination, management support, and alliance evaluation. The second dimension is alliance integration capability which incorporates relational capabilities, inter-organizational communication, relational capital, and project team effectiveness. The third is alliance learning capability which measures alliance experience, knowledge articulation, knowledge sharing, knowledge codification, internalization, and relationship learning. Although this study successfully develops a set of measurement items for alliance capabilities in university-industry R&D, further statistical analysis is required to test this scale. Conclusion: To date, quantitative measurement items for alliance capabilities in the context of university-industry R&D alliances are still at the infancy stage. Although the measurements are yet to be statistically analyzed, they can be used as a benchmark for future university-industry R&D alliances studies.

THE SCENARIO OF ACADEMIC STOMATHERAPY LEAGUES IN BRAZIL

The search for deepening in specific areas has grown substantially in professional training, in order to investigate and share knowledge, together with active participation in practical activities, in order to preserve the inseparability of the teaching-research-extension tripod1 . Through this perception, strategies were adopted that promote immersion and articulation of knowledge between theory and practice, through the creation of Academic Leagues (AL), which consist of organizations characterized by their own statute, exempt from profit, composed of academics and oriented by professionals who are members of certain institutions, in an extracurricular manner.

BERTtoCNN: Similarity-preserving enhanced knowledge distillation for stance detection.

In recent years, text sentiment analysis has attracted wide attention, and promoted the rise and development of stance detection research. The purpose of stance detection is to determine the author’s stance (favor or against) towards a specific target or proposition in the text. Pre-trained language models like BERT have been proven to perform well in this task. However, in many reality scenes, they are usually very expensive in computation, because such heavy models are difficult to implement with limited resources. To improve the efficiency while ensuring the performance, we propose a knowledge distillation model BERTtoCNN, which combines the classic distillation loss and similarity-preserving loss in a joint knowledge distillation framework. On the one hand, BERTtoCNN provides an efficient distillation process to train a novel ‘student’ CNN structure from a much larger ‘teacher’ language model BERT. On the other hand, based on the similarity-preserving loss function, BERTtoCNN guides the training of a student network, so that input pairs with similar (dissimilar) activation in the teacher network have similar (dissimilar) activation in the student network. We conduct experiments and test the proposed model on the open Chinese and English stance detection datasets. The experimental results show that our model outperforms the competitive baseline methods obviously.

Planning an interdisciplinary educational artifact: potentials, weaknesses and challenges

Currently, there are concerns about the fragmented teaching of science. Thus, interdisciplinary teaching is assumed as a trend, capable of presenting students with a vision of the whole, closer to reality. In this sense, this article seeks to describe the planning of an interdisciplinary artifact designed to promote the articulation of knowledge from the Natural Sciences. It is an excerpt from a master's research, completed last year. It is a descriptive research, guided by the theoretical-methodological assumptions of Design-based Research. Here, the description of the process is specifically presented with regard to the following steps: formation of the design group; identification and analysis of the problem in the school context; analysis and search for solutions; construction and organization of the educational artifact. As a result, the organization and planning of an interdisciplinary work on the Radiation and Radioactivity theme is presented, highlighting the potential, weaknesses and challenges of the process. It is believed that this work can help researchers and professors who aim to carry out interdisciplinary projects and promote the articulation of knowledge in their practices. Furthermore, as much as this work deals with the disciplines of Natural Sciences, the way it was carried out can be replicated for articulation of concepts referring to other areas of knowledge, or even for the articulation between different areas of knowledge.

Enablers of postacquisition joint knowledge creation: evidence from joint patenting in high-tech mergers and acquisitions

PurposeThis study aims to investigate three organizational-level factors (geographical proximity, technological similarity and organizational identity integration) that influence the likelihood of post-mergers and acquisition (M&A) joint knowledge creation between inventors from the target and acquiring firms.Design/methodology/approachAnalyzing post-M&A joint patent filing activities from 136 M&A deals in high-tech industries, the authors conduct a zero-inflated negative binomial regression analysis to estimate the likelihood of joint knowledge creation.FindingsThe results indicate that close geographical proximity and technological similarity are positively associated, whereas the integration of organizational identity is negatively associated with post-M&A joint knowledge creation.Practical implicationsManagers searching for an ideal acquisition target firm for joint production of new technologies or products should consider factors such as location, prior knowledge base and post-acquisition integration strategies.Originality/valueThe proposed model is a comprehensive framework that considers physical, cognitive and identity dimensions as antecedents of post-M&A joint knowledge creation. This study analyzes joint patenting activities to measure post-M&A joint knowledge creation between target and acquiring firm inventors.

Practice submissions – are doctoral regulations and policies responding to the needs of creative practice?

ABSTRACT Practice-based research is now widely accepted at doctoral level, and it is recognised that creative practice can be the mode, method, tool, object, subject and/or embodiment of research in the arts and humanities PhD. The growth of creative methods and arts-based methods also means that data is increasingly gathered through creative means in many social science and education doctorates. The doctoral contribution as thesis can therefore no longer be automatically assumed to be contained solely in a written text. This paper questions the extent to which research degree regulations and policies are reflecting and enabling the diversity of contemporary forms of knowledge articulation in practice-based research. Arising from my lived experience of supporting doctoral candidates to navigate regulations on the format and formatting of a submission, it draws on empirical research into research degree regulations at a number of universities in the United Kingdom, contextualised in relation to the literature. I reveal the assumptions and constraints embedded in regulatory practices and highlight ongoing concerns around the articulation and archiving of practice-based doctoral research.

Recommending Participants for Collaborative Merge Sessions

Development of large projects often involves parallel work performed in multiple branches. Eventually, these branches need to be reintegrated through a merge operation. During merge, conflicts may arise and developers need to communicate to reach consensus about the desired resolution. For this reason, including the right developers to a collaborative merge session is fundamental. However, this task can be difficult especially when many different developers have made significant changes on each branch over a large number of files. In this paper, we present TIPMerge, an approach designed to recommend participants for collaborative merge sessions. TIPMerge analyzes the project history and builds a ranked list of developers who are the most appropriate to integrate a pair of branches (Developer Ranking) by considering developers' changes in the branches, in the previous history, and in the dependencies among files across branches. Simply selecting the top developers in such a ranking is easy, but is not effective for collaborative merge sessions as the top developers may have overlapping knowledge. To support collaborative merge, TIPMerge employs optimization techniques to recommend developers with complementary knowledge (Team Recommendation) aiming to maximize joint knowledge coverage. Our results show a mean normalized improvement of 49.5% (median 50.4%) for the joint knowledge coverage with the optimization techniques for assembling teams of three developers for collaborative merge in comparison to choosing the top-3 developers in the ranked list.

To Burn or not to Burn? The History behind the Construction of a New Paradigm of Fire Management in Venezuela through Interculturality

The presence of a savanna-forest vegetation mosaic in the Gran Sabana, Canaima National Park (CNP), has been more than an academic controversy since the 1980s through to the 2010s in Venezuela. Scientists, Park administrators and officials from institutions devoted to protect the Caroní river basin within the limits of the Park, argued that the presence of savannas under a tropical rainfall regime that could support humid forests was due to the fire practices used by the local Pemón Indigenous communities. This misconception justified applying fire suppression policies, aimed at putting out all types of fires in CNP, especially the “compulsive burning” (thus so called by some scientists) by Pemón people in forest areas. This paper describes the initiatives, pursued for more than 20 years, to consolidate intercultural and participatory fire management in the CNP, Gran Sabana, which evolved as result of several participatory action-research projects coordinated by academics, and supported by national and regional public development institutions. The inclusion of Pemón Indigenous communities, firefighters, public officials, and academics in field research and joint experimentation, as well as in debates and dialogues on socio-ecological issues relevant to CNP, allowed the development of articulated knowledge and actions that were the foundations of a new paradigm of fire management and strategies for climate change mitigation and adaptation. Since 2015, these actions have been further extended to neighbouring countries of the Guiana Shield in the Northern Amazonia and other regions of Latin-America, with the collaboration and support of Venezuelan, British and other European institutions.

Creating collaboration for contentious projects on managed aquifer recharge: two cases from Finland

In Finland, community water supply has increasingly relied on natural groundwater and artificially recharged groundwater as the raw water source. Several managed aquifer recharge (MAR) projects have been co-created with involved parties and have proceeded well, while some cases have raised considerable resistance among the stakeholders. It seems that success or failure in MAR cooperation is related to management cultures and the ways in which various interests are taken into account, from the very beginning and throughout the process. Empirically, this paper builds on comparison between two conflictual case studies in Finland: one in the Tampere region and the other in the Turku region. The study analyses the major constraints of these projects through the lens of collaborative rationality, also drawing upon discourse analysis and negotiation theory. The material is gathered through thematic interviews of stakeholders, newspaper articles and a stakeholder workshop. The results indicate that conventional management approaches, drawing from expert-based instrumental rationality, were insufficient in both cases. The collaborative rationality framework suggests that legitimacy for the groundwater projects should be gained through joint knowledge production and inclusive multiparty interaction for creating options for collaboration. Both cases lacked the tools and know-how for authentic dialogue and collaboration. The emerging paradigm emphasizes more collaborative approaches for natural resources management and urban planning. While MAR projects operate inside these areas and are highly complex in nature, it is essential to embrace the emerging paradigm in order to promote MAR systems along with their huge potential.

Organization of the educational process on anti-doping support in a sports university

Main problem: the organization of anti-doping education of students assumes the presence of a specialized discipline "Fundamentals of anti-doping support" and the inclusion of anti-doping topics in classical disciplines. The formation of a number of competencies in the field of anti-doping education requires new interactive forms of organization of activities that take place in the development of both disciplines of the medical and biological cycle, as well as sports, pedagogical and social disciplines. A conscious perception of the industry problems associated with the use of doping in sports and the health of the nation provides knowledge of chemistry, biochemistry, physiology, sports medicine. Purpose: study of the organization of the educational process for anti-doping support in a sports university. Methods: today, the main methodological innovations are associated with the use of interactive learning technologies. The learning process, based on the use of interactive teaching methods, is organized taking into account the inclusion of all participants in the learning process. An educational communication environment is created, which is characterized by openness, interaction of participants, equality of their arguments, accumulation of joint knowledge, and the possibility of mutual evaluation. The organization of the educational process on anti-doping support for InEU students assumes the presence in the educational program of a specialized discipline "Fundamentals of anti-doping Support" and the introduction of components of anti-doping education in already established disciplines. Results and their significance: the organization of the educational process on anti-doping support for InEU students assumes the presence of a specialized discipline "Fundamentals of anti-doping Support" in Educaional Program and the inclusion of components of anti-doping education in the formed disciplines.

The Problem of Understanding in Russian Idealism of the Late 19<sup>th</sup> and Early 20<sup>th</sup> Centuries: V.S. Solovyov and S.N. Trubetskoy

The paper examines the doctrine of understanding in Russian idealism from the late 19th to early 20th century. The author discusses the main ontological and epistemological concepts in the philosophy of V.S. Solovyov and his follower S.N. Trubetskoy. The paper offers a historical and philosophical reconstruction of the concept of understanding based on the analysis of V.S. Solovyov's Lectures on Divine Humanity and S.N. Trubetskoy's work On the Nature of Human Consciousness. According to Solovyov, the study of understanding is possible if we presume an unconditional principle that premises being as the One. Based on this premise, it is demonstrated that the unconditional principle allows to know everything without having knowledge of each subject separately, and this, in turn, is understanding process. Trubetskoy's theory of understanding is premised on the concept of soborny (gathered) mind, which defines mind as joint knowledge about all spheres of existence. The article shows that the main position shared by the Russian thinkers is the unconditional principle serving as the basis for life and mind without which the existence of either the universal or the particular is impossible. The author concludes that such a principle is a cognitive relation to the world, which unconditionality is the condition for understanding the reality as indivisible. Understanding, as presented by V.S. Solovyov and S.N. Trubetskoy, is the true understanding of the unity of existence in all its manifestations, that is, a synthesis of the sensual, rational, moral, and aesthetic.

EDUCAÇÃO PROFISSIONAL E TECNOLÓGICA: desafios e perspectivas na formação docente

O presente ensaio tem por objetivo refletir a partir de uma contextualização histórica e legal da Educação Profissional e Tecnológica (EPT), sobre os desafios e as perspectivas na formação de seus profissionais. A metodologia empregada pautou-se em uma pesquisa de abordagem qualitativa, ancorada no método do materialismo histórico dialético e nos objetivos exploratórios e descritivos. Também utilizou-se da técnica bibliográfica narrativa, por meio de obras de autores e comentadores que examinam acerca da formação docente para atuar na EPT. Este ensaio apresenta uma estrutura que pode ser assim descrita: a) trata de uma breve contextualização histórica e legal da EPT no Brasil; b) reflete sobre os desafios e as perspectivas da formação de professores para a EPT; c) apresenta as considerações finais. Conclui-se que a EPT exige uma formação docente constituinte, sistemática e capaz de permitir a articulação e o desenvolvimento equilibrado de saberes gerais e técnicos, estes específicos de um campo da ciência.

A big data state of mind: Epistemological challenges to accountability and transparency in data-driven regulation

In a sense, the 2008 financial crisis was a crisis of theory. Regulators, banks, and financial markets all had encompassing theoretical models about how the economy worked, but they all failed to predict the looming crisis. As such, regulators increasingly turn to big data to understand banks' health. Despite the prominence of big data in society, its use in the public sector remains grossly understudied. This paper explores the regulatory use of big data in the case of the EU-wide banking stress test, a key regulatory indicator. The paper draws on interviews with supervisors at the European Central Bank (ECB), European Banking Authority (EBA) and National Bank of Belgium (NBB), as well as with consultants and risk directors in Belgian banks, to explain how big data-driven regulation affects the relationship between regulators and regulated entities. It draws particular attention to the epistemological component of using large data sets in decision-making: a big data state of mind. The article more specifically shows how the underlying epistemology, rather than simply the bigness of datasets, affects the relationship between regulators and regulated entities, and the regulatory process at large. The paper concludes that regulators' big data state of mind calls for new practical and legal guidelines regarding the validity of data-driven knowledge claims. Moreover, it shows how accountability based on descriptive transparency no longer makes sense in the ‘age of the algorithm’, suggesting a shift towards relational transparency and joint knowledge production.

Loops and Building Blocks: a Knowledge co-Production Framework for Equitable Urban Health

This paper sets out a structured process for the co-production of knowledge between researchers and societal partners and illustrates its application in an urban health equity project in Accra, Ghana. The main insight of this approach is that research and knowledge co-production is always partial, both in the sense of being incomplete, as well as being circumscribed by the interests of participating researchers and societal partners. A second insight is that project-bound societal engagement takes place in a broader context of public and policy debate. The approach to co-production described here is formed of three recursive processes: co-designing, co-analysing, and co-creating knowledge. These ‘co-production loops’ are themselves iterative, each representing a stage of knowledge production. Each loop is operationalized through a series of research and engagement practices, which we call building blocks. Building blocks are activities and interaction-based methods aimed at bringing together a range of participants involved in joint knowledge production. In practice, recursive iterations within loops may be limited due of constraints on time, resources, or attention. We suggest that co-production loops and building blocks are deployed flexibly.

Script concordance testing to understand the hypothesis processes of undergraduate nursing students - Multiple case study

Background: Albeit essential to clinical reasoning (CR), strategies for generating student nursing clinical hypotheses at the time of transition to professional practice are underdeveloped. While script concordance testing (SCT) has been shown to be a valid and reliable assessment tool for CR in nursing education, the thought processes including the hypothesis processes involved in choosing an answer is not examined.Methods: A multiple case study was used to understand the complex phenomenon of students’ hypothesis activation and confrontation with the combined use of SCT questions and the think-aloud method. Structured individual interviews were conducted.Results: A total of 18 students, nine first-year and nine third-year students participated in the study. The results show that the students demonstrate certain CR cognitive processes, including early representation of a clinical situation, semantic transformation of data, and hypothesis comparison.Conclusions: Results suggest promoting knowledge articulation aloud and the frequent use of micro-judgments to compare and differentiate hypotheses involving the uncertainty of clinical practice, which underpin learning in successive layers.

Potentials of blockchain technologies for supply chain collaboration: a conceptual framework

PurposeThe purpose of this study is to investigate the potentials of blockchain technologies (BC) for supply chain collaboration (SCC).Design/methodology/approachBuilding on a narrative literature review and analysis of seminal SCC research, BC characteristics are integrated into a conceptual framework consisting of seven key dimensions: information sharing, resource sharing, decision synchronization, goal congruence, incentive alignment, collaborative communication and joint knowledge creation. The relevance of each category is briefly assessed.FindingsBC technologies can impact collaboration between transaction partners in modern supply chains (SCs) by streamlining information sharing processes, by supporting decision and reward models and by strengthening communicative relationships with SC partners. BC promises important future capabilities in SCs by facilitating auditability, improving accountability, enhancing data and information transparency and improving trust in B2B relationships. The technology also promises to strengthen collaboration and to overcome vulnerabilities related to moral hazard and shortcomings found in legacy technologies.Research limitations/implicationsThe paper is mainly focused on the potentials of BC technologies on SCC as envisioned in the current academic literature. Hence, there is a need to validate the theoretical inferences with other approaches such as expert interviews and empirical tests. This study is of use to practitioners and decision-makers seeking to engage in BC-collaborative SC models.Originality/valueThe value of this paper lies in its call for an increased focus on the possibilities of BC technologies to support SCC. This study also contributes to the literature by filling the knowledge gap of how BC potentially impacts SC management.

Interfirm knowledge transfer in joint patenting: a technological distance analysis

In today’s highly competitive business environment, firms can ill afford the consequences of overlooking opportunities for innovation. An organization seeking to increase its existing knowledge base, in most cases, aims to identify and pursue useful knowledge available outside its borders. However, for companies interested in engaging in collaborative arrangements for innovations, systematic approaches for knowledge transfer may become a major challenge. In our research we focus on the diffusion of innovative knowledge that occurs during a joint knowledge development process. We present an empirical study spanning the period 2004 to 2018 which aims to explore the impact inter-organizational collaboration in the form of joint patenting has on the distance between partners’ technological bases. In addition, this paper looks into the change in technological distance when joint patenting occurs between different-country and same-country partners. The empirical results of the study suggest that engagement in joint patenting positively influences the technological proximity between partners and indicates a transmission and utilization of knowledge outside of firms’ existing knowledge base. Our findings contribute to knowledge related to innovation under inter?organizational collaboration and provides a basis for further theory development and testing.

Negotiating the Data Deluge on YouTube: Practices of Knowledge Appropriation and Articulated Ambiguity Around Visual Scenarios of Sea-Level Rise Futures

The present study aims at evaluating how YouTube users understand, negotiate and appropriate science-related knowledge on YouTube. It is informed by the qualitative analysis of post-video discussions around visual scenarios of sea-level rise (SLR) triggered by climate change. On the one hand, the SLR maps have an exemplary status as contemporary visualizations of climate change risks, beyond traditional image categories such as scientific or popular imagery. YouTube, on the other hand, is a convenient media environment to investigate the situated appropriation of such visual knowledge, considering its increasing relevance as a navigational platform to provide, search, consume and debate science-related information. The paper draws on media practice theory and operationalizes digital methods and qualitative coding informed by Grounded Theory. It characterizes a number of communicative practices of articulated knowledge appropriation regarding climate knowledge. This includes “locating impacts,” “demanding representation,” “envisioning further,” “debating future action,” “relativizing the information,” “challenging the reality of anthropogenic climate change,” “embedding popular narratives,” “attributing to politics,” and “insulting others.” The article then discusses broader questions posed by the comments and related to the appropriation and discursive negotiation of knowledge within online video-sharing platforms. Ambiguity is identified as a major feature within the practice of science-related information retrieval and knowledge appropriation on YouTube. This consideration then serves as an opportunity to reconsider the relationship between information credibility and knowledge appropriation in the age of the digital. Findings suggest that ambiguity of information can have a positive impact on problem definition, future imagination and the discursive negotiation of climate change.