Search results

Based on the strong feature representation ability of the convolutional neural network (CNN), generous object detection methods in remote sensing (RS) have been proposed one after another. However, due to the large variation in scale and the omission of relevant relationships between objects, there are still great challenges for object detection in RS. Most object detection methods fail to take the difficulties of detecting small and medium-sized objects and global context into account. Moreover, inference time and lightness are also major pain points in the field of RS.

To alleviate the aforementioned problems, this study proposes a novel method for object detection in RS, which is called lightweight object detection with a multi-receptive field and long-range dependency in RS images (MFLD). The multi-receptive field extraction (MRFE) and long-range dependency information extraction (LDIE) modules are put forward.

To concentrate on the variability of objects in RS, MRFE effectively expands the receptive field by a combination of atrous separable convolutions with different dilated rates. Considering the shortcomings of CNN in extracting global information, LDIE is designed to capture the relationships between objects. Extensive experiments over public datasets in RS images demonstrate that our MFLD method surpasses the state-of-the-art methods. Most of all, on the NWPU VHR-10 dataset, our MFLD method achieves 94.6% mean average precision with 4.08 M model volume.

This paper proposed a method called lightweight object detection with multi-receptive field and long-range dependency in RS images.

Research on the political connections of multinational enterprises’ (MNEs’) subsidiaries in emerging host countries has been growing. The purpose of this paper is to integrate institutional and resource dependence theories to argue that MNEs-subsidiaries are likely to develop fewer formal (i.e. board-level) political connections when operating in welfare-state monarchies as compared to in host countries with developmental-state democratic systems. Furthermore, this paper argues that MNE-subsidiaries develop formal political connections to a greater extent in industries where religion influences the development of products and services considerably. Finally, the extent of developing formal political connections varies by the scale of the MNEs’ investment (or subsidiary density) in the host market.

The paper tests its hypotheses on a sample of foreign-owned subsidiaries operating in Saudi Arabia and Egypt. The data was collected by combining information from Bureau Van Dijk’s Orbis database with company websites and other secondary sources. The final sample consisted of 156 observations – 70 MNEs-subsidiaries operating in Saudi Arabia, and 86 in Egypt.

The findings confirm that foreign subsidiaries are likely to develop fewer formal political connections in a welfare-state monarchy as compared to in a developmental-state democratic system. Furthermore, formal political connections are more significant in industries that are impacted by the influence of religion – such as the financial industry in Arab countries. Finally, the extent of using political connections varies by the scale of the MNEs’ investment in the host market – that is, with a greater scale of investment (or higher subsidiary density), formal political connections are greater.

The paper contributes theoretically by explaining that a combination of institutional heterogeneity and its associated resource dependence conditions between MNEs and host governments influence MNE-subsidiaries' political connections. The paper tests its hypotheses in an emerging Arab context, which is characterized by both autocratic and semi-democratic political settings, and which makes the integration of institutional and resource dependence theories useful in explaining how MNE-subsidiaries navigate local complexities in this region.

The present research aims to empirically test the “Barriers to abusive supervision model” to find how employee-related (core self-evaluations) and situational factors (perceived job dependency) make an employee trapped in the spiral of supervisory abuse. In addition, the work–family spillover lens is used to explain how employees' retaliation is targeted at their families in response to abuse from their bosses.

The current study has employed a three-wave longitudinal moderated mediation design and analysed data from 265 employees working in the hospitality industry of Pakistan.

The results of this study have shown that low core-self evaluations put employees in a spiral of supervisory abuse and they instil aggression towards their families. This association is further strengthened when employees are dependent on their job.

This study is one of the first to use the “Barriers to Abusive supervision” model to answer who and in which conditions tend to trap in the spiral of abuse and integrate the work-to-family interface model for elaborating the outcomes to the family domain.

Just-in-Time (JIT) arrival in the context of port calls can be used to reduce fuel and emissions to achieve environmental targets. The purpose of this paper is to study the implementation process of the Pre-booking Berth Allocation Policy (PBP) and analyze the effectiveness of this policy for the implementation of JIT in port calls.

The study deploys a single case study approach to empirically analyze port authority’s transition from a first-come-first-served (FCFS) arrival policy to the PBP. Observations, interviews and documents were used to collect data during 2020–2022. The analysis deployed the capability, opportunity, motivation and behavior model.

The transition from FCFS to PBP requires an inter-organizational approach, engaging external actors to manage diverse needs and preferences. This fosters effective transition and addresses conflicting interests. The PBP enables JIT arrival, enhancing operational and environmental performance, but faces barriers such as resource dependency and lack of trust. Information sharing capability among the actors, supported by Port Community Systems and adjusted operating rules, is crucial. Moreover, the PBP facilitates integration between sea and hinterland transportation, improving planning and efficiency across maritime transportation chains.

The single case study limits the generalizability of the findings.

Implementing the PBP is complex and demands careful planning from managers. Involving port call actors in the transition is helpful for port managers because they provide valuable feedback and highlight overlooked issues.

Five propositions are suggested to highlight the role of inter-organizational collaboration, information sharing and overcoming barriers such as resource dependency to successfully realize the benefits of JIT in maritime transportation chains.

This study examines the types, quality, and financial effects of explanations for non-compliance (NCEs) with corporate governance codes.

This study used content analysis to examine various types of NCEs and developed an NCE index (NCEI) to assess their quality and degree of informativeness. Static and dynamic multivariate panel data regression models were used to analyze the relationship between NCEI and firm performance (FP) of 38 non-financial listed Mauritian firms from 2009 to 2019.

Listed Mauritian firms do not provide explanations for all non-compliance, and the most common type of NCE is momentary deviation. The NCEI is 0.243, which implies that the overall quality of the NCEs is poor or uninformative. The NCEI varies according to the listing status and industry type. NCEI has a negative and insignificant relationship with both ROA and Tobin’s Q. The results are inconsistent with the agency, stakeholder, stewardship, and resource dependency theories. Sensitivity analysis indicated that the findings were robust.

Multiple theoretical frameworks offer a deeper understanding of corporate governance practices than a single theory does. A decline in the NCEI in 2019 indicates that the move from the “comply or explain” to the “apply or explain” principle does not necessarily result in enhancements in the degree of informativeness. Regulators should develop guidelines on how to disclose NCEs better. Investors appear to be more concerned about “comply/apply or perform” than the “comply/apply or explain” approach.

This study adds to the extant literature by providing new evidence on the types and quality of NCEs as well as their relationship with FP in emerging economies, where such studies are rare.

The purpose of this study is to investigate the influence of green finance on human well-being in China in the context of urbanization and aging population. It aims to explore the contributions of green finance in such demographic scenarios.

This study innovates and optimizes the calculation of the carbon intensity of human well-being (CIWB) index and strengthens the integrity of the assessment model for green finance development. It uses the serial multiple mediator model and moderation effect analysis to address the impact of green finance on human well-being in China on the provincial level from 2009 to 2020.

Green finance has a significant, positive and direct impact on human well-being. Simultaneously, it influences human well-being indirectly through three transmission channels. Urbanization and an ageing population are significant individual mediators through which green finance contributes to human well-being improvement. Notably, these two mediators also work together to transfer the promotional impact of green finance to human well-being.

The government can perfect the regulations to strengthen the market ecosystem to accelerate the development of green finance. Reforms on the administrative division to expand the size of cities with the implementation of ageing friendly development strategy is also necessary. Attracting incoming foreign direct investment in sustainable projects and adjusting public projects and trade activities to fulfil the sustainable principles are also regarded as essential.

The findings challenge traditional views on the impact of aging populations, highlighting the beneficial role of green finance in improving well-being amidst demographic changes. This offers a new perspective on economic and environmental sustainability in aging societies.

A multi-dimensional well-being indicator, CIWB and the serial multiple mediator model are used and direct and indirect impacts of green finance on human well-being is exhibited. It offers novel insights on the transmission channels behind, identifies the mediating role of urbanization and ageing population and offers empirical evidences with strong academic and policy implications.

Recent trends have shown the integration of Chinese word segmentation (CWS) and part-of-speech (POS) tagging to enhance syntactic and semantic parsing. However, the potential utility of hierarchical and structural information in these tasks remains underexplored. This study aims to leverage multiple external knowledge sources (e.g. syntactic and semantic features, lexicons) through various modules for the joint task.

We introduce a novel learning framework for the joint CWS and POS tagging task, utilizing graph convolutional networks (GCNs) to encode syntactic structure and semantic features. The framework also incorporates a pre-defined lexicon through a lexicon attention module. We evaluate our model on a range of public corpora, including CTB5, PKU and UD, the novel ZX dataset and the comprehensive CTB9 dataset.

Experimental results on these benchmark corpora demonstrate the effectiveness of our model in improving the performance of the joint task. Notably, we find that syntax information significantly enhances performance, while lexicon information helps mitigate the issue of out-of-vocabulary (OOV) words.

This study introduces a comprehensive approach to the joint CWS and POS tagging task by combining multiple features. Moreover, the proposed framework offers potential adaptability to other sequence labeling tasks, such as named entity recognition (NER).

Social media is booming in the digital age, and its rich availability provides many opportunities for companies to innovate across borders. In reality, how enterprises use social media to achieve cross-border innovation also faces important challenges such as breaking path dependency.

This paper explores how social media can facilitate cross-border innovation from the perspective of strategic capability, combined with the path dependency theory and attention-based view. Hierarchical regression analysis and bootstrap method are adopted to test the hypotheses based on survey data provided by 173 firms in China.

The findings show a positive relationship between social media strategic capability and cross-border innovation, with path dependency playing a mediating role. In addition, two internal and external contextual factors, namely customer embeddedness and competitive pressure, play moderating roles, with customer embeddedness negatively moderating the negative relationship between social media strategic capability and path dependency and competitive pressure negatively moderating the negative relationship between path dependency and cross-border innovation.

These findings provide not only new insights into social media and cross-border innovation but also theoretical guidance on how companies can effectively use social media in practice.

Drawing on a theoretical framework of sustainable career ecosystem theory, our paper aims to consider how domestic and international postgraduates can enhance their employability through participation in a landscape of practice.

The study employed an exploratory, longitudinal case study design to capture students' lived experiences on an 18-month Master of Professional Practice course at a higher education institution in New Zealand. The data collection procedure involved field note observations (months 1–4), a focus group (month 13) and narrative frames (months 16–18). The sample was domestic students from New Zealand (n = 2) and international students from Asia (n = 5).

One’s participation in multiple communities of practice represents their landscape of practice and a commitment to lifewide learning. Through participation in various communities of practice, domestic and international students can enhance their employability in three ways: (1) boundary encounters to develop social capital, (2) transcending contexts to enhance cultural capital, and (3) acknowledging the development of psychological capital and career agency.

Our work offers one of the earliest empirical validations of sustainable career ecosystem theory. Expressly, communities of practice represent various contexts whereby employability capital is developed over time. Additionally, the postgraduate students themselves are portrayed as interconnected and interdependent actors, presenting a novel framing of such dependencies at the micro-level of the ecosystem. The practical implications come from informing universities of the value of a landscape of practice to enhance the employability of domestic and international students in preparation for sustainable careers and to promote the sustainability of the career ecosystem.

Scientific impact is traditionally assessed with citation-based metrics. Recently, altmetric indices have been introduced to measure scientific impact both within academia and among the general public. However, little research has investigated the association between the linguistic features of research article titles and received online attention. To address this issue, the authors examined in the present study the relationship between a series of title features and altmetric attention scores.

The data included 8,658 titles of Science articles. The authors extracted six features from the title corpus (i.e. mean word length, lexical sophistication, lexical density, title length, syntactic dependency length and sentiment score). The authors performed Spearman’s rank analyses to analyze the correlations between these features and online impact. The authors then conducted a stepwise backward multiple regression to identify predictors for the articles' online impact.

The correlation analyses revealed weak but significant correlations between all six title features and the altmetric attention scores. The regression analysis showed that four linguistic features of titles (mean word length, lexical sophistication, title length and sentiment score) have modest predictive effects on the online impact of research articles.

In the internet era with the widespread use of social media and online platforms, it is becoming increasingly important for researchers to adapt to the changing context of research evaluation. This study identifies several linguistic features that deserve scholars’ attention in the writing of article titles. It also has practical implications for academic administrators and pedagogical implications for instructors of academic writing courses.