5916 research outputs found
Sort by
Benefits and challenges of implementing the common data environments in the construction phase of BIM-based projects
No full textThis research aims to identify, categorize, and prioritize the benefits and challenges of implementing common data environments (CDEs) in the construction phase of building information modeling (BIM)-based projects. For this purpose, a systematic literature review, semistructured interviews, and the Delphi method were conducted, and online sources of CDE providers were investigated. 54 benefits and 15 challenges were identified. Cronbach's alpha, standard deviation, interrater agreement level, and significance level statistics were calculated to analyze and validate Delphi method outputs. Results demonstrate that the most important benefit of implementing CDEs in the construction phase of BIM-based projects is access to accurate, reliable, up-to-date information. The most critical challenge is the exclusion of CDE planning in the front-end phase of the project. Prioritizing the benefits shows CDEs' full potential for the construction phase of BIM-based projects. Prioritizing the challenges demonstrates possible future works, requirements, and the areas that should be improved. This research provides a comprehensive point of view on the benefits and challenges of implementing CDEs in the construction phase of BIM-based projects. It analyzes the potential of CDEs by strengthening the findings from the systematic literature review with the SMEs' evaluation, comments, and feedback. This study provides valuable insights into implementing CDEs in the construction phase of BIM-based projects, defining their capabilities, status, and potential, and giving suggestions about how to deal with their challenges. This research shows how CDEs ensure streamlining information and communication flows in BIM-based projects by mitigating nonvalue-added activities resulting from informal, nonstandardized data exchanges, lack of proper data exchange protocols, loss of project data, and lack of efficient collaboration and communication between project team participants. This study reveals the importance of implementing CDEs in the construction phase of BIM-based projects to achieve adequate information and communication management that may prevent project data loss, informal, nonstandardized data exchanges, and inefficient collaboration between multidisciplinary teams
SDN-API-Sec: A conflict-free BC-based authorization for cross-domain SDNs
No full textSoftware-Defined Networking (SDN) is a novel paradigm that transforms network management into a centralized, flexible, and programmable task. However, this new paradigm adds new attack vectors severely targeting the decision making components in the SDN network (i.e., controller and applications). Therefore, there is a need to protect the Application Programming Interfaces (APIs) of SDN. In this work, we take advantage of the intrinsic characteristics of Blockchain (BC) technology to protect critical SDN resources. Our proposed system, SDN-API-Sec, provides a flexible conflict-free cross-domain access control method for SDN components through BC smart contracts. We also compare SDN-API-Sec with B-DAC
Bidirectional progressive neural networks with episodic return progress for emergent task sequencing and robotic skill transfer
No full textHuman brain and behavior provide a rich venue that can inspire novel control and learning methods for robotics. In an attempt to exemplify such a development by inspiring how humans acquire knowledge and transfer skills among tasks, we introduce a novel multi-task reinforcement learning framework named Episodic Return Progress with Bidirectional Progressive Neural Networks (ERP-BPNN). The proposed ERP-BPNN model 1) learns in a human-like interleaved manner by 2) autonomous task switching based on a novel intrinsic motivation signal and, in contrast to existing methods, 3) allows bidirectional skill transfer among tasks. ERP-BPNN is a general architecture applicable to several multi-task learning settings; in this paper, we present the details of its neural architecture and show its ability to enable effective learning and skill transfer among morphologically different robots in a reaching task. The developed Bidirectional Progressive Neural Network (BPNN) architecture enables bidirectional skill transfer without requiring incremental training and seamlessly integrates with online task arbitration. The task arbitration mechanism developed is based on soft Episodic Return progress (ERP), a novel intrinsic motivation (IM) signal. To evaluate our method, we use quantifiable robotics metrics such as 'expected distance to goal' and 'path straightness' in addition to the usual reward-based measure of episodic return common in reinforcement learning. With simulation experiments, we show that ERP-BPNN achieves faster cumulative convergence and improves performance in all metrics considered among morphologically different robots compared to the baselines. Overall, our method provides a human-inspired and efficient multi-task reinforcement learning approach with interleaved learning, making it highly suitable for lifelong learning applications.Japan Society for the Promotion of Science (JSPS) KAKENHI ; New Energy and Industrial Technology Development Organization (NEDO) ; INVERSEPublisher versio
The construction of reconciliation and social cohesion by Turkey’s Republican People’s Party: A critical discourse analysis of Kılıçdaroğlu’s Helalleşme call
No full textIn November 2021, Turkey's Republican People's Party (CHP) leader, Kilicdaroglu,1 1 shared a video via Twitter in which he made a call for helalle & scedil;me2 2 to Turkish society. While the repercussions of the call have been reflected differently within the political spectrum in Turkish society, this study investigates how social cohesion and reconciliation are envisaged and addressed by the main opposition party. It does this through a discursive analysis of Kilicdaroglu's speeches released since his helallesme call that builds upon theoretical discussions about reconciliation and intergroup forgiveness. Research on reconciliation and intergroup forgiveness literature associated with the Turkish case mainly revolves around the Turkish government's or Turkish nationalists' approach to ethnic minorities, such as the Kurdish question. However, studies investigating the stance of the main opposition party concerning the debate about fragmentation within Turkish society are very limited. Therefore, this study will aim to fill this gap. To thematically analyze the underlying messages of and motivation for the discourses, it will employ the discourse historical approach (DHA). In this way, the current study will contribute empirically to the relevant literature
İş hayatında çalışma biçimleri, kişisel tercihler ve öznel iyilik hali.
No full textTelework is a work arrangement that uses technology to work from home or any other location other than the office (Spreitzer et al., 2017). Many pieces of the literature showed that those who are more sociable and attracted to developing relationships with others (i.e., higher on sociability) tended to report lower ratings on indices of telework performance (O'Neill et al., 2009). In this study, we focused on the relationship between individual needs and subjective well-being and aimed to see if work arrangements play a moderator role. Also, with this study, we further explored how commuting convenience and commuting time affect subjective well-being in a big metropolitan city such as Istanbul. Data for the current study were collected from 203 (123 Females, 80 Males) white collar employees who work for several companies in Istanbul by Qualtrics. Employees' current work arrangements (Hybrid, face-to-face and remote), their number of remote working days, commuting time, their individual needs (the need for affiliation, the need for status and the need for achievement) and subjective well-being are measured. Bivariate correlations and hierarchical regression models were conducted to test the hypotheses. Results showed that work arrangements don't play a role as a moderator and there isn't any significant relationship between individual needs and subjective well-being. However, we found that commuting convenience predicts subjective well-being and higher commuting times decrease subjective well-being. The findings are discussed based on the previous literature, future directions and implications.Uzaktan çalışma, evden ya da ofis dışındaki herhangi bir yerden teknoloji kullanımıyla gerçekleştirilen bir çalışma düzenidir (Spreitzer, 2017). Önceki araştırmalar, sosyalleşme ihtiyacı yüksek olan kişilerin, uzaktan çalışma performanslarını diğer kişilere göre daha düşük olarak puanladıklarını göstermiştir (O'Neill, 2009). Bu çalışmada, bireysel ihtiyaçlar ile öznel iyilik hali arasındaki ilişkiye odaklanılarak, bu ilişkide çalışma düzeninin moderatör olarak rol alıp almadığı incelenmiştir. Ayrıca büyük bir metropol olan İstanbul'da yaşayan katılımcıların işe gidiş-gelişlerine ilişkin rahatlık algılarının ve işe gidiş-geliş ortalama sürelerinin öznel iyilik hallerini nasıl etkilediği incelenmiştir. Veriler İstanbul'da çeşitli firmalarda çalışan 203 (123 kadın, 80 erkek) beyaz yaka çalışandan Qualtrics sistemi üzerinden toplanmıştır. Çalışanların mevcut çalışma düzenleri (hibrit, yüz yüze, uzaktan), bir ay içinde uzaktan çalıştıkları gün sayıları, işe gidiş-geliş süreleri, işe gidiş gelişlerine ilişkin rahatlık algıları, bireysel ihtiyaç çeşitleri (ilişki ihtiyacı, statü ihtiyacı ve başarı ihtiyacı) ve öznel iyilik hallerine ilişkin ölçümler alınmıştır. İki değişkenli korelasyonlar ve hiyerarşik regresyon modelleri, hipotezleri test etmek için çalıştırılmıştır. Bulgular, çalışma düzenlerinin, bireysel ihtiyaçlar ve öznel iyilik hali ilişkisinde herhangi bir moderatör rolü oynamadığını göstermiştir. Bunun yanı sıra sonuçlar, çalışanların işe gidiş gelişlerine ilişkin rahatlık algılarının öznel iyilik halini yordadığını ve işe gidiş-geliş süresinin artmasının öznel iyilik halini azalttığını ortaya çıkarmıştır. Elde edilen sonuçlar geçmiş literatür ışığında incelenmiş ve gelecek araştırmalar için önerilerde bulunulmuştur
A game-theoretic analysis of the adoption of patient-generated health data
No full textPatient-generated health data (PGHD) have great potential to improve clinical outcomes. As providers consider whether and how to incorporate PGHD into their clinical workflows, platforms by Apple and Amazon stand to fundamentally alter the landscape. With the aim to examine the conditions under which providers would adopt PGHD and possibly sign on with a platform, we analyzed the incentives and optimal strategies of two healthcare providers, a monopoly platform, and consumers using stylized game-theoretic models and solve for potential equilibria. We found that consumer surplus always increased with PGHD adoption, but social welfare may drop. The larger provider had more incentive to adopt PGHD than the smaller provider, but these incentives were reversed in the case of platform adoption. Accordingly, the platform enrolled the smaller provider first and possibly both providers. The emergence of the platform raised provider surplus, potentially at the expense of the consumers, despite offering its service to them for free. These results illustrate the importance of economic incentives regarding whether and how PGHD could be incorporated into our current healthcare system.Publisher versio
Static and dynamic atom displacements in nanocrystalline powders from diffraction analysis at finite temperature
No full textWe evaluate the accuracy of x-ray diffraction analysis performed by Rietveld refinement in extracting the static and dynamic atomic displacement parameters of spherical gold nanocrystals with sizes of 30 nm and below and within a temperature interval of 10-300 K. A self-consistent computational workflow composed of real- and reciprocal- space analyses is designed for this purpose. We obtain average lattice parameters, static and thermal atomic displacements as well as their temperature and particle size dependence in an ideally monodispersed and realistic nanocrystalline powder generated by molecular dynamics simulations. Then, we compare crystallographic parameters obtained from real- and reciprocal- space analyses to evaluate the accuracy of diffraction analysis. We also extend the comparison to local variations of crystallographic parameters in the nanocrystals and show how Rietveld refinement relates to the average parameters of the core and surface regions separately. We show that 1) average lattice parameters obtained from Rietveld refinement are greater than the mean lattice parameter of small nanocrystals and converge to the core lattice parameter with increasing particle size, 2) Rietveld refinement significantly underestimates static atom displacements of surface atoms in small nanocrystals, 3) dynamic atomic displacements in nanocrystalline gold powders are reliably obtained within 5 % error by Rietveld refinement, provided that thermal diffuse scattering is adequately removed from the diffraction data, 4) the discrepancy between real- and reciprocal-space analysis increases for the smallest gold nanocrystal.TÜBİTA
Towards a unified approach for software architecture recovery
No full textLarge scale software systems are decomposed into gross-level components to reduce maintenance efforts. Software Architecture Recovery (SAR) approaches have been introduced to automatically cluster software modules for identifying modular components. In our previous work, we analyzed and hybridized various algorithms for clustering software modules. In addition to the employed clustering algorithm, SAR approaches are subject to many variations from multiple perspectives like the types of input and recovered artifacts. In this paper, we systematically review these variations. We aim at extending our work towards a unified approach for SAR in the future.TÜBİTA
Derin protein dil modellerini transformatörlerle birleştirerek rna ve protein modifikasyonlarini tahmin etmek ve analiz etmek.
No full textRecent work on language models has resulted in state-of-the-art performance on various language tasks. Among these, Bidirectional Encoder Representations from Transformers (BERT) has focused on contextualizing word embeddings to extract the context and semantics of the words. Besides, their protein-specific versions such as ProtBERT generated dynamic protein sequence embeddings which resulted in better performance for several bioinformatics tasks. On the other hand, Post-transcriptional 2'-O-methylation (Nm) RNA modification and a number of different protein post-translational modifications are prominent in cellular tasks and related to a number of diseases. The existing high-throughput experimental techniques take longer time to detect these modifications, and costly in exploring these functional processes. Here, to deeply understand the associated biological processes faster, we come up with two efficient methods: the first one is BERT2OME to infer 2'-O-methylation RNA modification sites from RNA sequences and the second one is DEEPPTM to predict protein post-translational modification (PTM) sites from protein sequences more efficiently. BERT2OME combines BERT-based model with convolutional neural networks (CNN) to infer the relationship between the modification sites and RNA sequence content. Unlike the methods proposed so far, BERT2OME assumes each given RNA sequence as a text and focuses on improving the modification prediction performance by integrating the pre-trained deep learning-based language model BERT. Additionally, our transformer-based approach could infer modification sites across multiple species. According to 5-fold cross-validation, human and mouse accuracies were 99.15% and 94.35% respectively. Similarly, ROC AUC scores were 0.99 and 0.94 for the same species. Detailed results show that BERT2OME reduces the time consumed in biological experiments and outperforms the existing approaches across different datasets and species over multiple metrics. Additionally, deep learning approaches such as 2D CNNs are more promising in learning BERT attributes than more conventional machine learning methods. Different than the current methods, DEEPPTM enhances the modification prediction performance by integrating specialized ProtBERT-based protein embeddings with attention-based vision transformers (ViT), and reveals the associations between different modification types and protein sequence content. Additionally, it can infer several different modifications over different species. Human and mouse ROC AUCs for predicting Succinylation modifications were 0.988 and 0.965 respectively, once 10-fold cross-validation is applied. Similarly, we have obtained 0.982, 0.955, and 0.953 ROC AUC scores on inferring ubiquitination, crotonylation, and glycation sites respectively. According to detailed computational experiments, DEEPPTM lessens the time spent in laboratory experiments while outperforming the competing methods as well as baselines on inferring all 4 modification sites. In our case, attention-based deep learning methods such as vision transformers look more favorable to learn from ProtBERT features than more traditional deep learning and machine learning techniques. Additionally, the protein-specific ProtBERT model is more effective than the original BERT embeddings for PTM prediction tasks.Dil modelleri üzerine yapılan son çalışmalar, çeşitli dil problemlerinde son derece iyi performans gösterdi. Örneğin, Transformatörler'de Çift Yönlü Kodlayıcı Gösterimleri (BERT), kelimelerin bağlamını ve anlamını çıkarma konusuna odaklanmıştır. ProtBERT gibi proteine özgü versiyonlar da, çeşitli biyoinformatik çalışmalarında çok iyi sonuçlar elde etmiştir. Bu çalışmaların yanısıra, transkripsiyon sonrası 2'-O-metilasyon (Nm) RNA modifikasyonu ve bir dizi farklı protein translasyon sonrası modifikasyonu, sadece hücresel görevlerde öne çıkmakla kalmayıp, canlılarda meydana gelen pekçok hastalıkla ilişkilendirilmiştir. Mevcut yüksek verimli deneysel tekniklerin bu değişiklikleri tespit etmesi hem uzun zaman almakta hem de çok fazla maaliyete gerektirmektedir. Bu alanlarda yapılacak ilgili biyolojik süreçleri daha hızlı hale getirmek ve ilgili konularda daha iyi sonuçlara ulaşmak için iki etkili model tasarladık; ilk olarak, RNA dizilerinden 2'-O-metilasyon RNA modifikasyon bölgelerini çıkarmak için kullanılacak olan BERT2OME ve ikinci olarak da protein dizilerinden protein translasyon sonrası modifikasyon (PTM) bölgelerini daha verimli bir şekilde tahmin etmek için DEEPPTM'dir. BERT2OME, modifikasyon bölgeleri ile RNA dizisi içeriği arasındaki ilişkiyi ortaya çıkarmak için BERT tabanlı modeli evrişimli sinir ağlarıyla (CNN) birleştirir. Şu ana kadar önerilen yöntemlerden farklı olarak BERT2OME, verilen her RNA dizisini bir metin olarak kabul eder ve önceden eğitilmiş derin öğrenme tabanlı dil modeli BERT'i entegre ederek modifikasyon tahmin performansını iyileştirmeye odaklanır. Transformatör tabanlı bu yaklaşımımız, birden fazla türün modifikasyon bölgelerini de ortaya çıkarabilmektedir. 5 katmanlı çapraz doğrulamaya göre insan ve farelerde doğru tahminleme başarısı sırasıyla 99.15% ve 94.35% olarak ölçüldü. Benzer şekilde ROC AUC skorları da aynı tür için 0.99 ve 0.94 olarak bulundu. Ayrıntılı sonuçlar, BERT2OME'nin biyolojik deneylerde harcanan süreyi azalttığını ve birden fazla ölçüm üzerinden farklı veri kümeleri ve türler genelinde mevcut yaklaşımlardan daha iyi performans gösterdiğini kanıtlamaktadır. Ek olarak, 2 boyutlu CNN'ler gibi derin öğrenme yaklaşımları, BERT özelliklerinin öğrenilmesinde daha geleneksel makine öğrenme yöntemlerine göre daha umut vericidir. Bir diğer modelimiz DEEPPTM, mevcut yöntemlerden farklı olarak, ProtBERT tabanlı protein yerleştirmelerini dikkat tabanlı görüntü transformatörleri (ViT) ile entegre ederek modifikasyon tahmin performansını arttırıp, farklı modifikasyon türleri ile protein dizisi içeriği arasındaki ilişkileri ortaya çıkarmıştır. Süksinilasyon modifikasyonlarını tahmin etmek için insan ve fare ROC AUC'leri, 10 katmanlı çapraz doğrulama uygulandığında sırasıyla 0.988 ve 0.965 sonuçlarını vermiştir. Benzer şekilde, ubikuitinasyon, krotonilasyon ve glikasyon bölgelerinin çıkarılmasında sırasıyla 0.982, 0.955 ve 0.953 ROC AUC skorları elde edilmiştir. DEEPPTM modelinin geliştirilmesinde, görüntü transformatörleri gibi dikkat odaklı derin öğrenme yöntemleri, ProtBERT özelliklerini öğrenme konusunda daha etkili olmuş ve daha geleneksel derin öğrenme ile makine öğrenme tekniklerine göre üstün bir performans sergilemiştir. Ek olarak proteine özgü ProtBERT modeli, PTM tahminlemesinde BERT modelinden daha etkili olmuştur
Incivility, mobbing, and abusive supervision: A tripartite scale development study
No full textUsing qualitative and quantitative methodologies, in three consecutive studies with employed samples, we developed measures of workplace incivility, mobbing, and abusive supervision sensitive to the nuances of a non-Western context (i.e., Turkiye). In Study 1, we first conducted 15 focus groups (N = 149), identified culture-specific and universal themes underlying the focal mistreatment types, and developed the initial scales. We then pilot-tested (N = 427) and refined the scales using exploratory factor analytic procedures. In Study 2, confirmatory factor analyses (N range = 456-524) and associations between the new scales and their widely used counterparts (N = 353) yielded evidence for the construct validity of the scales. Study 2 also involved the development of short forms of relatively long incivility and abusive supervision scales. In Study 3 (N = 482), we first examined the extent to which the three scales were operationally distinct. Second, we examined the scales' ability to predict burnout and organizational commitment. Results supported operational distinctiveness as well as the criterion-related validity of the scales. A dominance analysis revealed that the three scales had equivalent contributions in explaining the two outcome variables, further justifying their distinctiveness. We argue that the use of present scales is not necessarily restricted to the Turkish context and may prove useful more broadly in other neo-traditional contexts