Search results

This paper aims to provide an overview of key definitions related to ChatGPT, a public tool developed by OpenAI, and its underlying technology, Generative Pretrained Transformer (GPT).

This paper includes an interview with ChatGPT on its potential impact on academia and libraries. The interview discusses the benefits of ChatGPT such as improving search and discovery, reference and information services; cataloging and metadata generation; and content creation, as well as the ethical considerations that need to be taken into account, such as privacy and bias.

ChatGPT has considerable power to advance academia and librarianship in both anxiety-provoking and exciting new ways. However, it is important to consider how to use this technology responsibly and ethically, and to uncover how we, as professionals, can work alongside this technology to improve our work, rather than to abuse it or allow it to abuse us in the race to create new scholarly knowledge and educate future professionals.

This paper discusses the history and technology of GPT, including its generative pretrained transformer model, its ability to perform a wide range of language-based tasks and how ChatGPT uses this technology to function as a sophisticated chatbot.

Many strategies have been put forward for training deep network models, however, stacking of several layers of non-linearities typically results in poor propagation of gradients and activations. The purpose of this paper is to explore the use of two steps strategy where initial deep learning model is obtained first by unsupervised learning and then optimizing the initial deep learning model by fine tuning. A number of fine tuning algorithms are explored in this work for optimizing deep learning models. This includes proposing a new algorithm where Backpropagation with adaptive gain algorithm is integrated with Dropout technique and the authors evaluate its performance in the fine tuning of the pretrained deep network.

The parameters of deep neural networks are first learnt using greedy layer-wise unsupervised pretraining. The proposed technique is then used to perform supervised fine tuning of the deep neural network model. Extensive experimental study is performed to evaluate the performance of the proposed fine tuning technique on three benchmark data sets: USPS, Gisette and MNIST. The authors have tested the approach on varying size data sets which include randomly chosen training samples of size 20, 50, 70 and 100 percent from the original data set.

Through extensive experimental study, it is concluded that the two steps strategy and the proposed fine tuning technique significantly yield promising results in optimization of deep network models.

This paper proposes employing several algorithms for fine tuning of deep network model. A new approach that integrates adaptive gain Backpropagation (BP) algorithm with Dropout technique is proposed for fine tuning of deep networks. Evaluation and comparison of various algorithms proposed for fine tuning on three benchmark data sets is presented in the paper.

Recently, the double joint photographic experts group (JPEG) compression detection tasks have been paid much more attention in the field of Web image forensics. Although there are several useful methods proposed for double JPEG compression detection when the quantization matrices are different in the primary and secondary compression processes, it is still a difficult problem when the quantization matrices are the same. Moreover, those methods for the different or the same quantization matrices are implemented in independent ways. The paper aims to build a new unified framework for detecting the doubly JPEG compression.

First, the Y channel of JPEG images is cut into 8 × 8 nonoverlapping blocks, and two groups of features that characterize the artifacts caused by doubly JPEG compression with the same and the different quantization matrices are extracted on those blocks. Then, the Riemannian manifold learning is applied for dimensionality reduction while preserving the local intrinsic structure of the features. Finally, a deep stack autoencoder network with seven layers is designed to detect the doubly JPEG compression.

Experimental results with different quality factors have shown that the proposed approach performs much better than the state-of-the-art approaches.

To verify the integrity and authenticity of Web images, the research of double JPEG compression detection is increasingly paid more attentions.

This paper aims to propose a unified framework to detect the double JPEG compression in the scenario whether the quantization matrix is different or not, which means this approach can be applied in more practical Web forensics tasks.

Aiming at the problem that the transmission mechanism of the assembly error of the multi-stage rotor with saddle surface type is not clear, the purpose of this paper is to propose a deep belief network to realize the prediction of the coaxiality and perpendicularity of the multi-stage rotor.

First, the surface type of the aero-engine rotor is classified. The rotor surface profile sampling data is converted into image structure data, and a rotor surface type classifier based on convolutional neural network is established. Then, for the saddle surface rotor, a prediction model of coaxiality and perpendicularity based on deep belief network is established. To verify the effectiveness of the coaxiality and perpendicularity prediction method proposed in this paper, a multi-stage rotor coaxiality and perpendicularity assembly measurement experiment is carried out.

The results of this paper show that the accuracy rate of face type classification using convolutional neural network is 99%, which meets the requirements of subsequent assembly process. For the 80 sets of test samples, the average errors of the coaxiality and perpendicularity of the deep belief network prediction method are 0.1 and 1.6 µm, respectively.

Therefore, the method proposed in this paper can be used not only for rotor surface classification but also to guide the assembly of aero-engine multi-stage rotors.

Predicting highly cited papers can enable an evaluation of the potential of papers and the early detection and determination of academic achievement value. However, most highly cited paper prediction studies consider early citation information, so predicting highly cited papers by publication is challenging. Therefore, the authors propose a method for predicting early highly cited papers based on their own features.

This research analyzed academic papers published in the Journal of the Association for Computing Machinery (ACM) from 2000 to 2013. Five types of features were extracted: paper features, journal features, author features, reference features and semantic features. Subsequently, the authors applied a deep neural network (DNN), support vector machine (SVM), decision tree (DT) and logistic regression (LGR), and they predicted highly cited papers 1–3 years after publication.

Experimental results showed that early highly cited academic papers are predictable when they are first published. The authors’ prediction models showed considerable performance. This study further confirmed that the features of references and authors play an important role in predicting early highly cited papers. In addition, the proportion of high-quality journal references has a more significant impact on prediction.

Based on the available information at the time of publication, this study proposed an effective early highly cited paper prediction model. This study facilitates the early discovery and realization of the value of scientific and technological achievements.

This paper aims to examine the relationships between anthropomorphic cues (i.e. degrees of the humanized profile picture and naming) in artificial intelligence (AI) chatbots and business types (utilitarian-centered business vs hedonic-centered business) on consumers’ attitudes toward the AI chatbot and intentions to use the AI chatbot app and to accept the AI chatbot’s recommendation.

An online experiment with a 2 (humanized profile pictures: low [semihumanoid] vs high [full-humanoid]) × 2 (naming: Mary vs virtual assistant) × 2 (business types: utilitarian-centered business [bank] vs hedonic-centered business [café]) between-subjects design (N = 520 Mturk samples) was used.

The results of this study show significant main effects of anthropomorphic cues (i.e. degrees of profile picture and naming) in AI chatbots and three-way interactions among humanized profile pictures, naming and business types (utilitarian-centered business vs hedonic-centered business) on consumers’ attitudes toward the AI chatbot, intentions to use the AI chatbot app and intentions to accept the AI chatbot’s recommendation. This indicates that the high level of anthropomorphism generates more positive attitudes toward the AI chatbot and intentions to use the AI chatbot app and to accept the AI chatbot’s recommendation in the hedonic-centered business condition. Moreover, the mediated role of parasocial interaction occurs in this relationship.

This study is the original endeavor to examine the moderating role of business types influencing the effect of anthropomorphism on consumers’ responses, while existing literature overweighted the value of anthropomorphism in AI chatbots without considering the variation of businesses.

This work aims at proposing a novel Internet of Things (IoT)-based and cloud-assisted monitoring architecture for smart manufacturing systems able to evaluate their overall status and detect eventual anomalies occurring into the production. A novel artificial intelligence (AI) based technique, able to identify the specific anomalous event and the related risk classification for possible intervention, is hence proposed.

The proposed solution is a five-layer scalable and modular platform in Industry 5.0 perspective, where the crucial layer is the Cloud Cyber one. This embeds a novel anomaly detection solution, designed by leveraging control charts, autoencoders (AE) long short-term memory (LSTM) and Fuzzy Inference System (FIS). The proper combination of these methods allows, not only detecting the products defects, but also recognizing their causalities.

The proposed architecture, experimentally validated on a manufacturing system involved into the production of a solar thermal high-vacuum flat panel, provides to human operators information about anomalous events, where they occur, and crucial information about their risk levels.

Thanks to the abnormal risk panel; human operators and business managers are able, not only of remotely visualizing the real-time status of each production parameter, but also to properly face with the eventual anomalous events, only when necessary. This is especially relevant in an emergency situation, such as the COVID-19 pandemic.

The monitoring platform is one of the first attempts in leading modern manufacturing systems toward the Industry 5.0 concept. Indeed, it combines human strengths, IoT technology on machines, cloud-based solutions with AI and zero detect manufacturing strategies in a unified framework so to detect causalities in complex dynamic systems by enabling the possibility of products’ waste avoidance.

The purpose of this study is to develop a hybrid algorithm for segmenting tumor from ultrasound images of the liver.

After collecting the ultrasound images, contrast-limited adaptive histogram equalization approach (CLAHE) is applied as preprocessing, in order to enhance the visual quality of the images that helps in better segmentation. Then, adaptively regularized kernel-based fuzzy C means (ARKFCM) is used to segment tumor from the enhanced image along with local ternary pattern combined with selective level set approaches.

The proposed segmentation algorithm precisely segments the tumor portions from the enhanced images with lower computation cost. The proposed segmentation algorithm is compared with the existing algorithms and ground truth values in terms of Jaccard coefficient, dice coefficient, precision, Matthews correlation coefficient, f-score and accuracy. The experimental analysis shows that the proposed algorithm achieved 99.18% of accuracy and 92.17% of f-score value, which is better than the existing algorithms.

From the experimental analysis, the proposed ARKFCM with enhanced level set algorithm obtained better performance in ultrasound liver tumor segmentation related to graph-based algorithm. However, the proposed algorithm showed 3.11% improvement in dice coefficient compared to graph-based algorithm.

The image preprocessing is carried out using CLAHE algorithm. The preprocessed image is segmented by employing selective level set model and Local Ternary Pattern in ARKFCM algorithm. In this research, the proposed algorithm has advantages such as independence of clustering parameters, robustness in preserving the image details and optimal in finding the threshold value that effectively reduces the computational cost.

This paper aims to manage the word gap in information retrieval (IR) especially for long documents belonging to specific domains. In fact, with the continuous growth of text data that modern IR systems have to manage, existing solutions are needed to efficiently find the best set of documents for a given request. The words used to describe a query can differ from those used in related documents. Despite meaning closeness, nonoverlapping words are challenging for IR systems. This word gap becomes significant for long documents from specific domains.

To generate new words for a document, a deep learning (DL) masked language model is used to infer related words. Used DL models are pretrained on massive text data and carry common or specific domain knowledge to propose a better document representation.

The authors evaluate the approach of this study on specific IR domains with long documents to show the genericity of the proposed model and achieve encouraging results.

In this paper, to the best of the authors’ knowledge, an original unsupervised and modular IR system based on recent DL methods is introduced.

The purpose of this study is to analyze the intelligent semisupervised fault diagnosis method of aeroengine.

A semisupervised fault diagnosis method based on denoising autoencoder (DAE) and deep belief network (DBN) is proposed for aeroengine. Multiple state parameters of aeroengine with long time series are processed to form high-dimensional fault samples and corresponding fault types are taken as sample labels. DAE is applied for unsupervised learning of fault samples, so as to achieve denoised dimension-reduction features. Subsequently, the extracted features and sample labels are put into DBN for supervised learning. Thus, the semisupervised fault diagnosis of aeroengine can be achieved by the combination of unsupervised learning and supervised learning.

The JT9D aeroengine data set and simulated aeroengine data set are applied to test the effectiveness of the proposed method. The result shows that the semisupervised fault diagnosis method of aeroengine based on DAE and DBN has great robustness and can maintain high accuracy of fault diagnosis under noise interference. Compared with other traditional models and separate deep learning model, the proposed method also has lower error and higher accuracy of fault diagnosis.

Multiple state parameters with long time series are processed to form high-dimensional fault samples. As a typical unsupervised learning, DAE is used to denoise the fault samples and extract dimension-reduction features for future deep learning. Based on supervised learning, DBN is applied to process the extracted features and fault diagnosis of aeroengine with multiple state parameters can be achieved through the pretraining and reverse fine-tuning of restricted Boltzmann machines.