Forthcoming and Online First Articles
International Journal of Information and Communication Technology
Forthcoming articles have been peer-reviewed and accepted for publication but are pending final changes, are not yet published and may not appear here in their final order of publication until they are assigned to issues. Therefore, the content conforms to our standards but the presentation (e.g. typesetting and proof-reading) is not necessarily up to the Inderscience standard. Additionally, titles, authors, abstracts and keywords may change before publication. Articles will not be published until the final proofs are validated by their authors.
Forthcoming articles must be purchased for the purposes of research, teaching and private study only. These articles can be cited using the expression "in press". For example: Smith, J. (in press). Article Title. Journal Title.
Articles marked with this shopping trolley icon are available for purchase - click on the icon to send an email request to purchase.
Online First articles are published online here, before they appear in a journal issue. Online First articles are fully citeable, complete with a DOI. They can be cited, read, and downloaded. Online First articles are published as Open Access (OA) articles to make the latest research available as early as possible.
Articles marked with this Open Access icon are Online First articles. They are freely available and openly accessible to all without any restriction except the ones stated in their respective CC licenses.Register for our alerting service, which notifies you by email when new issues are published online.
International Journal of Information and Communication Technology (26 papers in press) Regular Issues
Abstract: The increasing expansion of digital media content presents a significant difficulty in the field of information processing since it becomes difficult to effectively create accurate and succinct summaries from large news sources. Especially with complex and multimodal news content, traditional news summary generating techniques are sometimes difficult to consider the information coverage; the impact is restricted. This work so suggests, based on the transformer model, an automatic summarising method for digital media news. While decreasing the development of repetitive content via redundancy penalty factors, sentence vector augmentation and keyword guiding techniques help to more precisely capture the relevant information in the news. The strategy suggested in this work much beats the conventional summary generating model in the ROUGE series of metrics. New technical solutions and value references for automated processing and intelligent generation of digital media news are presented by this work. Keywords: digital media news; automatic summarisation; transformer model; multimodal information; redundancy penalty; optimisation strategy. DOI: 10.1504/IJICT.2025.10071981
Abstract: This paper proposes an innovative solution based on the improved seagull optimisation algorithm (ISOA) for the key technical challenges of art sculpture style recognition in digital protection of cultural heritage. Through the design of quantum-chaos hybrid initialisation strategy and dynamic nonlinear parameter system, it breaks through the bottleneck of early convergence of traditional optimisation algorithms in high-dimensional feature space, and combines the 3D differential geometric feature enhancement model and multimodal data fusion technology to construct an intelligent recognition framework with deformation robustness. Experiments show that the algorithm achieves 96.5% recognition accuracy on the dataset, which is 8.2% higher than the mainstream model. It provides a new generation of technical paradigm for the protection and intelligent identification of art treasures. Keywords: improved seagull optimisation algorithm; ISOA; 3D differential geometric features; multimodal data fusion; digital preservation of cultural heritage. DOI: 10.1504/IJICT.2025.10071982
Abstract: Acoustic noise recognition of ancient buildings is crucial for the protection and study of ancient buildings, but the traditional methods have problems such as insufficient feature extraction and weak generalisation ability in complex scenes, which are ineffective for noise recognition. Therefore, this paper proposes an acoustic noise recognition method for ancient buildings based on space time joint processing and deep learning, which utilises space time joint processing to pre-process acoustic signals and extract effective features, and then classifies and recognises them through a deep learning model. Experiments show that the method shows excellent performance in terms of recognition accuracy and robustness, providing new ideas and effective means for the recognition of acoustic noise of ancient buildings, which helps to better protect and study ancient buildings. Keywords: noise recognition; spatiotemporal joint processing; deep learning; acoustic signals; ancient architecture. DOI: 10.1504/IJICT.2025.10071983
Abstract: To address traditional graph neural networks (GNNs) neglect of word-order features and sensitivity to adversarial noise, this study proposes DGCL-TC, a text classification model integrating dual-graph fusion and adaptive contrastive learning. The framework leverages bidirectional encoder representations from transformers (BERT) to encode contextual semantics and constructs dual graphs capturing local and global text structures. A learnable augmentation module dynamically generates contrastive views via node dropout and attribute masking, optimising representations through cross-view consistency. A gated graph attention network fuses topology-aware graph features with BERT embeddings, balancing structural and sequential cues. Evaluations on benchmark datasets confirm that DGCL-TC significantly outperforms baseline methods in accuracy and robustness, particularly under adversarial perturbations and sparse data conditions. The model advances text classification by unifying semantic, structural, and noise-resistant representation learning. Keywords: text classification; graph neural network; GNNs; contrastive learning; BERT. DOI: 10.1504/IJICT.2025.10071984
Abstract: To address teaching semantic gap issues in image sample learning for vocational education evaluation, this paper first applies factor domain theory to the teaching semantic embedding domain. Based on the relationships among semantics, it studies conjunction and reduction of factors, as well as the expansion and contraction of the factor domain. The enhanced factor space approach is then utilised in vocational education evaluation. Visual features are extracted using the residual network (ResNet101), and a generative adversarial network (GAN) is trained to produce more realistic picture characteristics. By combining teaching attributes and noise, the generator outputs picture characteristics which are then combined with class marks to train the categoriser, thereby completing the classification of teaching evaluation images. Experimental results reveal that the offered model achieves a classification accuracy of 92.5%, effectively helping to enhance the quality of higher vocational education. Keywords: vocational education; factor space mathematical theory; machine classification learning; ResNet101 model; generative adversarial network; GAN. DOI: 10.1504/IJICT.2025.10071985
Abstract: With the rapid development of big data technology, traditional gradient descent algorithms face problems such as low computational efficiency, slow convergence speed, and uneven resource allocation. This article proposes a collaborative framework that integrates dynamic resource scheduling and adaptive gradient descent optimisation for distributed machine learning scenarios in big data environments. Firstly, an asynchronous gradient descent algorithm based on hierarchical batch sampling (HB-ASGD) was designed, which dynamically adjusts the local batch size and global synchronisation frequency to balance the load differences between computing nodes and reduce communication overhead. Secondly, the resource aware elastic scheduling (RAES) model is introduced to dynamically predict task computation using reinforcement learning, and combined with containerisation technology to achieve fine-grained allocation of CPU/GPU resources, prioritising the protection of computing resources for critical iterative tasks. The experiment shows that this study effectively solves the efficiency bottleneck problem in massive data iteration. Keywords: big data; gradient descent algorithm; resource scheduling; reinforcement learning; resource aware elastic scheduling; RAES. DOI: 10.1504/IJICT.2025.10071986
Abstract: Violin fingering techniques change rapidly, making it difficult to correct incorrect fingering in real time during instruction. To address this issue, this paper first employs a multi-head attention mechanism (MAM) and multi-scale dilated convolutional neural networks (DCNN) for hand fingering motion capture. Since hand movement occurs during performance, an augmented reality (AR)-based hand pose estimation module is designed. The pose from orthography and scaling with iterations (POSIT) algorithm, optimised using the Gauss-Newton method, is used to estimate relatively precise camera-based fingering poses. Finally, cosine similarity is used to compare virtual and real finger techniques, and corrections are made based on the features of the target finger technique to improve teaching effectiveness. Experimental results show that the proposed algorithm achieves a correction accuracy rate 3.66%13.13% higher than the baseline algorithm, laying a foundation for improving violin finger technique instruction. Keywords: violin fingering instruction; motion capture; augmented reality; multi-scale dilated convolution; POSIT algorithm. DOI: 10.1504/IJICT.2025.10071987
Abstract: The academic anxiety of English as a foreign language (EFL) learners has progressively become a major determinant of academic performance. Therefore, how best to forecast and evaluate the influence of these emotions on academic performance has become a hot issue of research in academia. This work suggests an academic anxiety emotion analysis based on graph neural network (GNN). This work uses the GNN model to investigate learners anxiety feelings in depth by processing multimodal data on their emotional traits, therefore verifying the accuracy of the approach in forecasting academic anxiety emotions. According to the experimental results, the suggested approach has tremendous generalisation capacity and application possibilities and greatly beats conventional machine learning methods in many respects. This paper offers fresh perspectives on the academic anxiety and theoretical support for the design of emotional control techniques in education of EFL learners. Keywords: graph neural network; GNN; academic anxiety; EFL learners; multimodal data; sentiment prediction; sentiment analysis. DOI: 10.1504/IJICT.2025.10071988
Abstract: City parks are increasingly crucial in improving inhabitants quality of life as well as other factors given the fast worldwide urbanisation. But as urban population density rises, the classic landscape design of urban parks has proved challenging to satisfy the ever varied needs of people. This work presents a multimodal data fusion-based landscape perception analysis and design optimisation method for urban parks, so building a comprehensive landscape perception model that can effectively capture the multidimensional traits of park environments and their effect on visitor experience, so solving these problems. Both tests reveal that multimodal data fusion greatly increases the comprehensiveness and accuracy of landscape perception. Although this study has achieved great progress, there are still certain constraints including the restricted capacity to react to dynamic data in real time; so, future studies will concentrate on addressing these issues and optimising the research approach. Keywords: multimodal data fusion; urban park; landscape perception; intelligent optimisation. DOI: 10.1504/IJICT.2025.10071989
Abstract: Social learning platforms provide a dynamic interactive and individualised learning experience as artificial intelligence in education is increasingly used. Though they have data privacy and AI security issues, these services gather much user data. Releasing user data such as voice calls, text messages, social connections, and learning preferences could cause privacy issues. This paper presents FL-KGPrivacyEdu, a federated knowledge graph model combining federated learning (FL) and knowledge graph (KG) to address the problem of behavioural modelling and privacy preservation in social English learning. The methodology examines social learning across platforms and terminals while safeguarding data. Experimental validation demonstrates that the model is successful and superior in social English learning behaviour analysis in accuracy, recall, and F1 score. This study also looks at how model convergence, a major model training reference, is affected by learning rate modification in worldwide rounds. Keywords: federated learning; FL; knowledge graph; KG; federated knowledge graph; data privacy protection; AI security. DOI: 10.1504/IJICT.2025.10071990
Abstract: Intending to the issue of low fidelity of images generated by existing fashion design methods, this paper firstly segments the original garment images semantically based on VGG and Unet, encodes the target garment part images into different part features, and obtains the appearance flow field of the target garment parts. Then the target garment parts are deformed according to the appearance flow field of the garment parts, and the features of each garment part are fused. Finally, the garment images are generated in light of generative adversarial network (GAN) and diffusion model, and the degrees of freedom are restricted by human posture control module and region degree discriminator to enhance the local fine-grainedness of the garment images. The experimental results show that the structural similarity index (SSIM) of the proposed method is 0.895, and the generated results are clearer and more realistic. Keywords: fashion design; automated generation; deep learning; generative adversarial network; GAN; diffusion model. DOI: 10.1504/IJICT.2025.10071991
Abstract: With the extensive use of electronic components in contemporary industry, fault diagnosis technology is more important in preserving equipment operation and improving output. Conventional fault diagnosis techniques limit their application in complicated fault situations by means of cross-domain feature extraction, which suffers limitations. This work thus suggests an electronic component fault diagnosis model called Cross-DeepContrastNet, which combines cross-domain feature extraction with deep contrastive learning and uses a series of training strategies to effectively extract discriminative features from many sources and types of data and acquire accurate fault diagnosis. Cross-DeepContrastNet beats conventional techniques in several respects, according to different experimental findings. Finally, further paths of investigation are suggested to solve the constraints of the use of the model in actual industries. Keywords: electronic components; fault diagnosis; cross-domain features; deep contrastive learning. DOI: 10.1504/IJICT.2025.10072071
Abstract: Owing to the defects of semantic segmentation of deep convolution network and the confusion and multi-scale problems of landslides in remote sensing images, large-scale spatial separation convolution kernel and multiscale fusion semantic segmentation network is proposed. By using a large spatially separable convolution and channel attention mechanism on the encoder, the landslide image is extracted with large-scale information, which ensures the accurate extraction of landslide edge information; A skip connection is adopted between the encoder and the decoder to recover the context loss caused by the down sampling of the encoder; At the same time, the atrous spatial pyramid pooling (ASPP) module is applied to extract and fuse multi-scale features, so as to further improve the performance. The experimental results show that the segmentation effect of the proposed network on landslide dataset is better than FCN, SegNet, U-Net, DeeplabV3+ and other semantic segmentation methods, and it also verifies that the network has good landslide recognition ability in medium and high vegetation coverage areas. Experimental results demonstrate that the proposed network significantly outperforms existing semantic segmentation methods such as FCN, SegNet, U-Net, and DeepLabV3+ on landslide datasets, and exhibits strong landslide recognition capabilities in areas with medium to high vegetation coverage. Keywords: landslide; semantic segmentation; attention mechanism; deep learning; receptive field; atrous spatial pyramid pooling; ASPP. DOI: 10.1504/IJICT.2025.10072177
Abstract: With the integration of computational intelligence (CI) in music education, the traditional teaching paradigms have been transformed, and personalised and adaptive learning experiences have been offered. This paper describes a methodology for making decisions on music education, which is based on artificial intelligence (AI), machine learning (ML) and data analytics. The AI-powered CI frameworks provide tailored instructional content, optimised practice regimes, and instant feedback through analytics of students abilities, preferences, and evolution. In this study, a critical review of existing CI techniques in music education has taken place intending to investigate their effectiveness in the areas of skill acquisition, engagement, and curriculum design. Furthermore, to increase the levels of personalisation in learning pathways, a novel decision-making model is proposed that ensures students receive instruction tailored to their abilities and goals. The experimental validation of the proposed methodology was carried out through case studies, demonstrating its effectiveness in improving student outcomes and learning processes within educational contexts. The studys conclusions reinforce the idea that the power of CI can transform music pedagogy from educational to Artificial Intelligence. The results indicate that AI-powered education is the way to develop learning opportunities in future programs. Keywords: computational intelligence; music education; personalised learning; decision-making models; artificial intelligence. DOI: 10.1504/IJICT.2025.10072178
Abstract: There are growing intersections between classical music performance, its tradition, acoustic fidelity, and advanced computational technologies, creating immersive multimedia concert experiences. We discuss how classical concert design is advanced by integrating real-time audio processing, machine learning models, generative visual engines, and audience interaction frameworks. A modular architecture was developed and tested through two case studies (Beethoven 360 and Bach Rewired), demonstrating two separate use cases of artificial intelligence, real-time rendering, and biofeedback systems. Sensor-based audience input has been used to learn about the adaptive performance elements, and musical structure was mapped to visual output using deep learning models. Results showed that audience immersion increased from an average of 5.4 to 6.8, while system latency was less than 25 ms. The findings suggest that computational multimedia systems might improve the classical concert experience by bringing new ways for emotional expression, structural clarity, or participatory design while maintaining the integrity of the repertoire. Keywords: multimedia concert design; computational music systems; artificial intelligence in music; real-time audio-visual synchronisation; generative visuals; audience interaction; machine learning in performance arts. DOI: 10.1504/IJICT.2025.10072179
Abstract: In this paper, an early warning system of cigarette process quality combined with intelligent sensing technology is proposed to improve the quality of cigarette process production. A PCA multi-block modelling algorithm based on autoencoder feature extraction is proposed to extract autoencoder features from each sub-block, and the statistics of all sub-blocks are fused by Bayesian inference to make the monitoring results more intuitive. Compared with the traditional PCA and AE-PCA detection methods, the AE-MPCA algorithm proposed in this paper improves the abnormality detection accuracy of the drum leaf drying production process, and realises the accurate alarm of quality abnormalities, thus providing technical support for the early warning of subsequent cigarette process quality. In the subsequent process of cigarette process quality control, the application scope of intelligent sensing technology can be further improved to promote the effect of cigarette process quality control. Keywords: intelligent perception; cigarettes; process quality; early warning. DOI: 10.1504/IJICT.2025.10071852
Abstract: This study explores the integration of machine learning (ML) and adaptive learning technologies in enhancing English teaching effectiveness in vocational colleges. A comprehensive dataset was collected from student interactions and feedback to evaluate engagement levels and learning outcomes. Text-based features such as TF-IDF, POS tagging, and Word2Vec embeddings were extracted and analysed using traditional ML and deep learning models including SVM, decision tree, naive Bayes, LSTM, and RNN. The hybrid CNN+ViT model achieved the highest classification accuracy of 92.7%, demonstrating the effectiveness of integrating machine learning for improving English teaching strategies. These findings suggest a data-driven path for optimising personalised instruction in English language education. Keywords: English teaching; vocational colleges; machine learning; adaptive learning; data-driven education. DOI: 10.1504/IJICT.2025.10071905
Abstract: The integration of deep learning algorithms in art classification has revolutionised the way artistic styles are identified and analysed. This study explores the application of neural networks - particularly convolutional neural networks (CNNs), generative adversarial networks (GANs) and vision transformers (ViTs) - in distinguishing and classifying various forms of art, including abstract, realism, impressionism, and digital art. By leveraging large datasets, these models can identify stylistic features with high accuracy. The paper compares the performance of different models and highlights the challenges of training on heterogeneous art databases, such as data imbalance and complex feature extraction. Results show the effectiveness of hybrid architectures like CNN + ViT, and potential future applications include museum curation, style transfer, and computational creativity. This research underlines the evolving role of AI in bridging technology and art. Keywords: deep learning; art classification; neural networks; style recognition; computational creativity. DOI: 10.1504/IJICT.2025.10071894
Abstract: In regional economic planning, accurate forecasting and efficient resource allocation are vital for informed decision-making by both government and private sectors. The article titled 'Dynamic path transformer network for regional economic forecasting and resource allocation' presents a novel deep learning-based approach that leverages transformer architecture to enhance forecasting precision and optimise the allocation of resources. Central to this study is the dynamic path transformer network (DPTN), which effectively captures complex spatial-temporal economic data through attention mechanisms that dynamically weigh economic indicators. This design allows the model to adapt to changing economic conditions and deliver more accurate predictions than traditional statistical or machine learning models. The study benchmarks DPTN against conventional approaches and demonstrates its superior performance in predictive accuracy and resource management. Moreover, the paper explores the broader implications for policy formulation and strategic planning, while also addressing key challenges such as data limitations, computational demands, and interpretability. Keywords: economic forecasting; deep learning; transformer network; resource allocation; spatial-temporal analysis. DOI: 10.1504/IJICT.2025.10071796
Abstract: The professional development of vocational college teachers is crucial for enhancing teaching quality and student outcomes. This paper presents an AI-driven framework using natural language processing (NLP) to support vocational educators' professional growth. The framework evaluates teaching effectiveness, identifies skill gaps, and provides customised feedback for continuous improvement. NLP techniques were applied to analyse the teaching content, feedback, and communication patterns of 100 teachers and 2,000 students over a 12-month period. The framework's success was evaluated using important measures: 1) teaching performance improved by 30% in evaluation accuracy; 2) student engagement increased by 20% in satisfaction with a 2% error margin; 3) skills in specific subjects grew by 35%; 4) feedback delivery became 45% faster. Results indicate substantial improvements in teaching performance, student satisfaction, and academic outcomes, along with optimal feedback delivery. This study demonstrates the potential for AI-driven NLP frameworks to revolutionise vocational education. Keywords: artificial intelligence; vocational education; teacher development; natural language processing; skill enhancement. DOI: 10.1504/IJICT.2025.10071906
Abstract: The integration of artificial intelligence (AI) in the design of rural landscape strategies is transforming traditional practices of sustainable development and spatial planning. This research paper thoroughly analyses how AI-based models - such as machine learning algorithms, geographic information systems (GIS), and deep learning techniques - facilitate decision-making in rural land-use planning, environmental conservation, and resource management. These technologies are increasingly employed to analyse large datasets, predict land-use changes, and optimise strategic interventions. AI fosters more efficient and adaptive planning by handling complex policy decisions with data-driven insights. The study evaluates the effectiveness, challenges, and future possibilities of AI-based systems, emphasising outcomes such as environmental stability, community well-being, and drought mitigation. A collaborative approach involving AI experts, environmental planners, and policymakers is essential for ethically and contextually relevant implementation. Keywords: artificial intelligence; rural landscape; strategy design; machine learning; sustainable planning; geographic information systems; GIS. DOI: 10.1504/IJICT.2025.10071797
Abstract: The distribution of teaching data varies among database courses, and traditional methods are often difficult to deal with such domain differences effectively. For this reason, this paper firstly utilises BERT model for embedding learning of teaching feedback text, and then extracts local and global features of the text through convolutional neural network (CNN) and long short-term memory (LSTM) network respectively, and enhances the text features through the attention mechanism. On this basis, the domain adaptive transfer learning algorithm is adopted to achieve the characteristic distribution migration alignment of the text source topic and objective topic, and minimise the scoring difference between different classifiers through consistency constraints, so as to assess the teaching quality more accurately. Simulation results show that the classification accuracy of the offered method is 94.39%, which demonstrates a substantial enhancement over the benchmark method. Keywords: database curriculum; teaching quality evaluation; BERT model; domain adaptation; transfer learning. DOI: 10.1504/IJICT.2025.10071895
Abstract: Deep learning and digital image processing powered by artificial intelligence are now influencing art education. With AI, artists can now experiment with new styles and effects, thanks to CNNs, GANs and NST. Tasks such as edge detection, segmentation and super-resolution give rise to helpful approaches in creative learning. AI-assisted art is represented by platforms such as DeepDream and RunwayML. While AI offers fast and original feedback to improve learning, many are worried about who should get credit for the results, ethics and the loss of traditional abilities. We must deal with problems such as dataset bias, copyright and having too much trust in AI. By being careful with AI, it can connect rather than conflict with conventional art, while aiming for ethics, diverse sets and blended ways of teaching. Keywords: AI in art education; deep learning in creativity; digital image processing; generative adversarial networks; GANs; neural style transfer; NST; ethical AI in art. DOI: 10.1504/IJICT.2025.10071872
Abstract: The assessment of marketing efficacy has grown more complicated and varied with the virtual economy's fast expansion. The conventional approaches of measuring marketing effectiveness are challenging to fit to the particular requirements of the virtual economy setting. Aiming to thoroughly evaluate marketing activities in virtual economy using quantitative and qualitative criteria including customer lifetime value (CLV), this paper offers a fuzzy comprehensive evaluation (FCE)-based marketing effectiveness evaluation model for virtual economy, i.e., GV-EFCE. The validity and superiority of the GV-EFCE model are verified by experiments run on two datasets. In many virtual economy situations, the experimental findings reveal that the GV-EFCE model surpasses conventional approaches. Thus, this work offers a fresh concept for assessing marketing efficacy in the domain of virtual economy and offers a useful guide for next studies. Keywords: virtual economy; marketing effectiveness evaluation; fuzzy comprehensive evaluation; FCE; customer lifetime value; CLV. DOI: 10.1504/IJICT.2025.10071980
Abstract: The study examines state-of-the-art artificial intelligence (AI) methodologies aimed at developing sports image classification as it affects multimedia management as well as recommendation algorithms and sport data analysis capabilities. The sports industry is witnessing unprecedented growth, fuelled by advancements in technology, and the exponential rise of digital content. The vast quantity of sports-related media requires critical management for improved accessibility for user engagement capabilities. AI brings transformative automation capabilities through its ability to tackle these sorts of tasks. Deep learning applications show outstanding performance for resolving intricate classification challenges. This research developed a sports image classification framework using deep neural networks (DNNs) and analysed two pre-trained models ResNet-50 and MobileNet for performance comparisons. The DNN model demonstrated outstanding performance metrics through 98% accuracy which matched its precision and recall and F1-scores. DNN proved the most suitable solution when compared to pre-trained models ResNet-50 and MobileNet. Keywords: artificial intelligence; sports classification; game; deep neural network; DNN; feature extraction. DOI: 10.1504/IJICT.2025.10071786
Abstract: In this paper, a novel approach is given to evaluate the effect of chorus size on the quality of art music performance using the hesitant bipolar fuzzy multi-criteria decision-making (HBFS-MCDM) framework. The method accounts for positive and negative expert evaluations under conditions of hesitancy concerning five performance criteria: tonal balance, articulation precision, dynamic range, audience cohesion, and emotional expression. To aggregate, expert conductors and musicologists evaluated a hesitant bipolar fuzzy number, and then a weighted decision model was proposed. Results show limited evidence of balanced and consistent performance among all evaluated dimensions across various ensemble sizes. The HBFS-MCDM approach offers a robust and expressive tool for supporting decisions in professional chorals and for actionable insights on ensemble configuration. Keywords: choral performance evaluation; chorus size; hesitant bipolar fuzzy sets; multi-criteria decision-making; MCDM; art music analysis; subjective judgment modelling. DOI: 10.1504/IJICT.2025.10071787 |
Open Access
Keep up-to-date
-
Our Blog -
Join us on Bluesky -
Follow us on X -
Visit us on Facebook - Our Newsletter (subscribe for free)
- RSS Feeds
- New issue alerts
- Inderscience is a member of publishing organisations including:
-
-
