Search results

Fine-tuning pre-trained language models (PLMs), e.g. SciBERT, generally require large numbers of annotated data to achieve state-of-the-art performance on a range of NLP tasks in the scientific domain. However, obtaining fine-tuning data for scientific NLP tasks is still challenging and expensive. In this paper, the authors propose the mix prompt tuning (MPT), which is a semi-supervised method aiming to alleviate the dependence on annotated data and improve the performance of multi-granularity academic function recognition tasks.

Specifically, the proposed method provides multi-perspective representations by combining manually designed prompt templates with automatically learned continuous prompt templates to help the given academic function recognition task take full advantage of knowledge in PLMs. Based on these prompt templates and the fine-tuned PLM, a large number of pseudo labels are assigned to the unlabelled examples. Finally, the authors further fine-tune the PLM using the pseudo training set. The authors evaluate the method on three academic function recognition tasks of different granularity including the citation function, the abstract sentence function and the keyword function, with data sets from the computer science domain and the biomedical domain.

Extensive experiments demonstrate the effectiveness of the method and statistically significant improvements against strong baselines. In particular, it achieves an average increase of 5% in Macro-F1 score compared with fine-tuning, and 6% in Macro-F1 score compared with other semi-supervised methods under low-resource settings.

In addition, MPT is a general method that can be easily applied to other low-resource scientific classification tasks.

Purpose: The primary purpose is to discuss the productivity and digitalisation interaction at the theoretical level, analyse the productivity and digitalisation differences between the European Union (EU)-14 and EU-13 countries, and evaluate the digitalisation impact on the manufacturing sector labour productivity of the EU countries.

Need for study: The average added value created per capita in new EU countries (EU-13) is one-third lower than in old EU countries (EU-14). To increase productivity, manufacturing companies must adapt to modern trends and take advantage of industrial digitisation opportunities. Digitisation can improve production efficiency, reduce costs, and improve product quality, allowing continuous monitoring and analysis of production data, enabling informed decisions and faster problem-solving.

Methodology: Analysis of scientific literature, comparing viewpoints, insights, and conclusions. The empirical study includes calculating rates of change of indicators, differences between EU-14 and EU-13, and structural analysis. The impact of digitisation on the productivity of EU countries is studied by creating a correlation matrix and using regression analysis: ordinary least square models.

Findings: EU-13 countries are behind EU-14 in labour productivity and manufacturing digitalisation. Digitalisation positively impacts productivity per employee. A faster increase in digitisation, industrial robot use, and e-commerce sales could significantly increase productivity in EU-13, reducing productivity differences between countries.

Practical implications: This study highlights the need for policy promoting digitisation innovation, particularly in EU-13 countries, to be implemented by both national and EU-based economic development and regional and cohesion institutions.

This study aims to propose a cloud platform architecture considering information sharing based on blockchain to realize the security and convenience of enterprise information sharing in the automotive supply chain.

A bilateral matching model considering enterprises information contribution stimulates information sharing and improves the efficiency and quality of supply and demand matching. Three smart contracts are used to complete the information sharing process and match supply and demand in the automotive supply chain.

The system is tested on the local Ganache private chain, and the decentralized web page is designed based on the architecture prototype.

Solve the problem of information island in automobile supply chain.

As technology aspect of agriculture becomes more and more important with the time to increase agricultural productivity in a sustainable and smart way, agriculture practices become more interdisciplinary. Furthermore, agricultural practices are affected by urban and rural planning enabling urban and rural farming. Architecture, engineering and construction (AEC) industry can support food security through the integration of agricultural practices and technologies into the built environment, its interior design, and greenhouses supporting urban and rural farming. Based on the literature review, this chapter aims to investigate ways for enhancing AEC industry’s and its professionals’ contribution to food security and sustainable agricultural practices. This chapter highlights roles of the AEC industry in enhancing food security and sustainable agricultural practices. This chapter emphasizes the importance of undergraduate and graduate curriculums of future AEC industry professionals (e.g., architects, interior architects, civil engineers) to equip them with the skills and knowledge of sustainable agricultural practices and technologies integrated greenhouses, built environment and indoor environment, and interior design. For this reason, agricultural policies need to cover food security-related interdisciplinary education and training (e.g., renewable energy-based agriculture integrated built environment) of AEC industry professionals. Agricultural policies need to be designed with the contribution of and considering AEC industry professionals as they are among the main stakeholders of food security and renewable energy-based agriculture-integrated built environment. Furthermore, this chapter highlights how AEC industry, in compliance with United Nations Sustainable Development Goals and countries sustainable and resilient development plans, can contribute to food security and sustainability. This chapter can be beneficial to all stakeholders of the sustainable agricultural practices.

This study aims to comprehensively examine the transitions in household livelihood strategies within rural China, including the underlying processes, drivers and outcomes.

This study uses two waves (2010 and 2018) of longitudinal data from the China Family Panel Studies (CFPS), employing latent cluster analysis, regression models and cumulative distribution function within a dynamic household livelihood strategy framework.

The results show that (1) households’ livelihood strategies can be categorized into four distinct types, i.e. agricultural dominated, agricultural dominated with non-agricultural supplementation, non-agricultural dominated with agricultural supplementation and employment oriented. (2) During 2010–2018, approximately 60% of households underwent transitions in their livelihoods, encompassing both upward and downward trajectories, with a prevalence of upward transitions. (3) Various forms of livelihood capital significantly contribute to upward transitions, while the availability of land resources and exposure to shocks impede the potential for upward mobility. (4) The transition towards non-agricultural livelihood strategies by households leads to a notable enhancement in their livelihood welfare.

In the context of urbanization, industrialization and globalization, rural areas in China are undergoing a gradual socioeconomic transformation, which has also led to changes in rural households’ livelihood strategies. Nevertheless, a dearth of empirical investigation exists regarding the dynamics of rural households’ livelihood strategies, the determinants behind such transitions and the resulting outcomes on their livelihoods. A comprehensive understanding of livelihood transitions can provide valuable insights for policymakers in their endeavors to promote rural revitalization in China.

Based on the nationwide representative datasets in China, it examines the micro-level livelihood transitions of rural households within the broader context of socioeconomic transformation that presents both opportunities and challenges, as well as vulnerable contexts, shaped by various government policies. This exploration would offer valuable theoretical and empirical evidence to advance our understanding of the process, driver and outcome of rural households’ livelihood transition in developing countries.

The main objective of this article is to visualize the structure and trends of pro-environmental behavior (PEB) between 1999 and 2023 through mapping and in-depth analysis. The aim is to analyze PEB, which has received considerable academic attention in recent years due to its key role in the conservation of the environment and the protection of local communities in tourist destinations. This paper provides an important summary of the recent research that has explored the role that tourists have in protecting the environment through PEB.

This study presents a visual analysis of 2005 scholarly articles between the years 1999 and 2023 related to PEB. Using the knowledge mapping based on VOSviewer it presents the current status of research, which includes the analysis of citation analysis, co-citation analysis, co-citation network and longitudinal analysis.

PEB is an emerging topic due to its relevance to protecting the environment in the context of travel. The citation and co-citation analysis show the relevance of the behavior of tourists with regard to protecting the environment. The co-word analysis highlights the current significance of research concerning green hotels and the destination image of environmentally responsible destinations.

This study sheds light on the current research progress of PEB in the context of tourism through a comprehensive analysis (citation, co-citation and co-word). In addition, we provide theories and factors that have been previously used to study PEB in the context of tourism. The findings contribute to a broad and diverse understanding of the concept of PEB, which can provide important insights for policymakers in formulating management strategies and policies aimed at reducing environmental impacts in destinations.

With the global market for advanced technology-driven customer service set to soar, understanding the complicated relationship between advanced technology and customer purchase behaviour is paramount. While prior research has touched upon the impact of technology on purchase processes in some aspects, this study investigates the specific features of advanced technology that shape customer purchase intention in greater depth. By investigating when and under what conditions customers choose advanced technology-based purchases, this research sheds light on the evolving landscape of consumer decision-making and it seeks to quantify the transformative power of advanced technology in driving customer purchase intentions.

Small islands are particularly vulnerable to environmental impacts, as multiple environmental as well as socio-economic changes are impacting their local communities and especially the most vulnerable segments of their population. Information and communication technologies (ICTs) are increasingly viewed as an opportunity for small islands to mitigate and adapt to climate change. ICT may help to monitor short-term and long-term climate trends, raise awareness, help protect the environment and reduce carbon emissions. Though the ICT sector has been recognised as crucial in ensuring sustainable development, it is also important to address its potential adverse impacts like energy consumption, electronic waste generation and digital inequality among others. The ICT-environment link is thus rather complex. While there is extensive literature on the ICT-climate change nexus, the evidence remains mixed. The evidence on small island economies is rather scant. The objective of this chapter is to investigate into the ICT and environment linkage for small islands taking on board the specificities of island economies. The Panel Vector Error Correction Model (PVECM) is used on 38 small islands over a period 2000–2020, and the long-run results show that higher use of ICT has resulted in lower carbon emissions.

The purpose of the present study is to synthesize and organize the existing literature on sustainable supply chain management (SSCM) challenges and potential solutions to overcome these challenges. The four-step research method has been used to collect and analyze pertinent literature, define the unit of analysis, select the classification context, collect publications, and evaluate the material. The study found 10 prevalent SSCM challenges and 18 potential solutions to overcome these challenges. By implementing these solutions, organizations can implement SSCM practices and contribute to achieving various Sustainable Development Goals (SDGs). The study significantly contributes to stakeholder theory, triple bottom-line theory, and resource-based view theory. The current study provides insights to managers working in supply chain management on SSCM implementation. Furthermore, the study also has practical implications for academicians and policymakers. This study is the first of its kind to amalgamate the SSCM challenges and solutions to overcome these challenges in a single framework by reviewing the literature.

Product returns are a major challenge for e-businesses as they involve huge logistical and operational costs. Therefore, it becomes crucial to predict returns in advance. This study aims to evaluate different genres of classifiers for product return chance prediction, and further optimizes the best performing model.

An e-commerce data set having categorical type attributes has been used for this study. Feature selection based on chi-square provides a selective features-set which is used as inputs for model building. Predictive models are attempted using individual classifiers, ensemble models and deep neural networks. For performance evaluation, 75:25 train/test split and 10-fold cross-validation strategies are used. To improve the predictability of the best performing classifier, hyperparameter tuning is performed using different optimization methods such as, random search, grid search, Bayesian approach and evolutionary models (genetic algorithm, differential evolution and particle swarm optimization).

A comparison of F1-scores revealed that the Bayesian approach outperformed all other optimization approaches in terms of accuracy. The predictability of the Bayesian-optimized model is further compared with that of other classifiers using experimental analysis. The Bayesian-optimized XGBoost model possessed superior performance, with accuracies of 77.80% and 70.35% for holdout and 10-fold cross-validation methods, respectively.

Given the anonymized data, the effects of individual attributes on outcomes could not be investigated in detail. The Bayesian-optimized predictive model may be used in decision support systems, enabling real-time prediction of returns and the implementation of preventive measures.

There are very few reported studies on predicting the chance of order return in e-businesses. To the best of the authors’ knowledge, this study is the first to compare different optimization methods and classifiers, demonstrating the superiority of the Bayesian-optimized XGBoost classification model for returns prediction.