scholarly journals Large Scale Subject Category Classification of Scholarly Papers With Deep Attentive Neural Networks

2021 ◽  
Vol 5 ◽  
Author(s):  
Bharath Kandimalla ◽  
Shaurya Rohatgi ◽  
Jian Wu ◽  
C. Lee Giles
Keyword(s):  
Neural Networks ◽  
Domain Knowledge ◽  
Large Scale ◽  
Knowledge Domain ◽  
Scientific Publications ◽  
Individual Subject ◽  
Subject Category ◽  
Sentence Level ◽  
Subject Categories

Subject categories of scholarly papers generally refer to the knowledge domain(s) to which the papers belong, examples being computer science or physics. Subject category classification is a prerequisite for bibliometric studies, organizing scientific publications for domain knowledge extraction, and facilitating faceted searches for digital library search engines. Unfortunately, many academic papers do not have such information as part of their metadata. Most existing methods for solving this task focus on unsupervised learning that often relies on citation networks. However, a complete list of papers citing the current paper may not be readily available. In particular, new papers that have few or no citations cannot be classified using such methods. Here, we propose a deep attentive neural network (DANN) that classifies scholarly papers using only their abstracts. The network is trained using nine million abstracts from Web of Science (WoS). We also use the WoS schema that covers 104 subject categories. The proposed network consists of two bi-directional recurrent neural networks followed by an attention layer. We compare our model against baselines by varying the architecture and text representation. Our best model achieves micro-F1 measure of 0.76 with F1 of individual subject categories ranging from 0.50 to 0.95. The results showed the importance of retraining word embedding models to maximize the vocabulary overlap and the effectiveness of the attention mechanism. The combination of word vectors with TFIDF outperforms character and sentence level embedding models. We discuss imbalanced samples and overlapping categories and suggest possible strategies for mitigation. We also determine the subject category distribution in CiteSeerX by classifying a random sample of one million academic papers.

Domain Knowledge Embedding Regularization Neural Networks for Workload Prediction and Analysis in Cloud Computing

2018 ◽  
Vol 11 (4) ◽  
pp. 137-154 ◽  
Author(s):  
Lei Li ◽  
Min Feng ◽  
Lianwen Jin ◽  
Shenjin Chen ◽  
Lihong Ma ◽  
...  
Keyword(s):  
Neural Networks ◽  
Cloud Computing ◽  
Artificial Neural Networks ◽  
Cost Effectiveness ◽  
Domain Knowledge ◽  
Large Scale ◽  
Workload Prediction ◽  
Platform As A Service ◽  
Resource Cost ◽  
Artificial Neural

Online services are now commonly deployed via cloud computing based on Infrastructure as a Service (IaaS) to Platform-as-a-Service (PaaS) and Software-as-a-Service (SaaS). However, workload is not constant over time, so guaranteeing the quality of service (QoS) and resource cost-effectiveness, which is determined by on-demand workload resource requirements, is a challenging issue. In this article, the authors propose a neural network-based-method termed domain knowledge embedding regularization neural networks (DKRNN) for large-scale workload prediction. Based on analyzing the statistical properties of a real large-scale workload, domain knowledge, which provides extended information about workload changes, is embedded into artificial neural networks (ANN) for linear regression to improve prediction accuracy. Furthermore, the regularization with noisy is combined to improve the generalization ability of artificial neural networks. The experiments demonstrate that the model can achieve more accuracy of workload prediction, provide more adaptive resource for higher resource cost effectiveness and have less impact on the QoS.

Exploring the knowledge domain of earthquake prediction based on bibliometric analysis and text mining

2021 ◽  
pp. 1-17
Author(s):  
Xiaorong He
Keyword(s):  
Text Mining ◽  
Bibliometric Analysis ◽  
Earthquake Prediction ◽  
Citation Index ◽  
Knowledge Domain ◽  
Scientific Publications ◽  
Main Research ◽  
Detection Techniques ◽  
Subject Categories ◽  
Earthquake Research

Earthquake prediction is one of the important themes of earthquake research, and it is also a very difficult scientific problem in the world. In this study, a bibliometric analysis is conducted on the scientific publications about earthquake prediction indexed in SCIE (Science Citation Index Expanded) and SSCI (Social Sciences Citation Index) databases during the past two decades (1998–2017). The subject categories, annual and journal distributions, leading countries/regions and institutions are investigated in this field. The main research topics are identified through text mining method. The research trends are explored by keyword co-occurrence analysis and bursting keywords detection techniques. The results of this study are helpful for scholars in this field to find the knowledge structure and important participants. It is also helpful for scholars to seize the current research hotspots and future development trends in this field.

scholarly journals Hybrid Representation of Sensor Data for the Classification of Driving Behaviour

2021 ◽  
Vol 11 (18) ◽  
pp. 8574
Author(s):  
Michalis Savelonas ◽  
Ioannis Vernikos ◽  
Dimitris Mantzekis ◽  
Evaggelos Spyrou ◽  
Athanasia Tsakiri ◽  
...  
Keyword(s):  
Neural Networks ◽  
Time Series ◽  
Event Detection ◽  
Domain Knowledge ◽  
Short Term Memory ◽  
Main Idea ◽  
Sensor Data ◽  
Support Vector ◽  
Driving Behaviour

Monitoring driving behaviour is important in controlling driving risk, fuel consumption, and CO2 emissions. Recent advances in machine learning, which include several variants of convolutional neural networks (CNNs), and recurrent neural networks (RNNs), such as long short-term memory (LSTM) and gated recurrent unit (GRU) networks, could be valuable for the development of objective and efficient computational tools in this direction. The main idea in this work is to complement data-driven classification of driving behaviour with rules derived from domain knowledge. In this light, we present a hybrid representation approach, which employs NN-based time-series encoding and rule-guided event detection. Histograms derived from the output of these two components are concatenated, normalized, and used to train a standard support vector machine (SVM). For the NN-based component, CNN-based, LSTM-based, and GRU-based variants are investigated. The CNN-based variant uses image-like representations of sensor measurements, whereas the RNN-based variants (LSTM and GRU) directly process sensor measurements in the form of time-series. Experimental evaluation on three datasets leads to the conclusion that the proposed approach outperforms a state-of-the-art camera-based approaches in distinguishing between normal and aggressive driving behaviour without using data derived from a camera. Moreover, it is demonstrated that both NN-guided time-series encoding and rule-guided event detection contribute to overall classification accuracy.

Domain Knowledge Embedding Regularization Neural Networks for Workload Prediction and Analysis in Cloud Computing

2021 ◽  
pp. 1158-1176
Author(s):  
Lei Li ◽  
Min Feng ◽  
Lianwen Jin ◽  
Shenjin Chen ◽  
Lihong Ma ◽  
...  
Keyword(s):  
Neural Networks ◽  
Cloud Computing ◽  
Artificial Neural Networks ◽  
Cost Effectiveness ◽  
Domain Knowledge ◽  
Large Scale ◽  
Workload Prediction ◽  
Resource Requirements ◽  
Resource Cost ◽  
Artificial Neural

Online services are now commonly deployed via cloud computing based on Infrastructure as a Service (IaaS) to Platform-as-a-Service (PaaS) and Software-as-a-Service (SaaS). However, workload is not constant over time, so guaranteeing the quality of service (QoS) and resource cost-effectiveness, which is determined by on-demand workload resource requirements, is a challenging issue. In this article, the authors propose a neural network-based-method termed domain knowledge embedding regularization neural networks (DKRNN) for large-scale workload prediction. Based on analyzing the statistical properties of a real large-scale workload, domain knowledge, which provides extended information about workload changes, is embedded into artificial neural networks (ANN) for linear regression to improve prediction accuracy. Furthermore, the regularization with noisy is combined to improve the generalization ability of artificial neural networks. The experiments demonstrate that the model can achieve more accuracy of workload prediction, provide more adaptive resource for higher resource cost effectiveness and have less impact on the QoS.

scholarly journals Graph Convolutional Networks by Architecture Search for PolSAR Image Classification

2021 ◽  
Vol 13 (7) ◽  
pp. 1404
Author(s):  
Hongying Liu ◽  
Derong Xu ◽  
Tianwen Zhu ◽  
Fanhua Shang ◽  
Yuanyuan Liu ◽  
...  
Keyword(s):  
Neural Networks ◽  
Large Scale ◽  
Spatial Relations ◽  
Classification Problem ◽  
Search Method ◽  
Sample Distribution ◽  
Convolutional Networks ◽  
Training Samples ◽  
Graph Neural Networks

Classification of polarimetric synthetic aperture radar (PolSAR) images has achieved good results due to the excellent fitting ability of neural networks with a large number of training samples. However, the performance of most convolutional neural networks (CNNs) degrades dramatically when only a few labeled training samples are available. As one well-known class of semi-supervised learning methods, graph convolutional networks (GCNs) have gained much attention recently to address the classification problem with only a few labeled samples. As the number of layers grows in the network, the parameters dramatically increase. It is challenging to determine an optimal architecture manually. In this paper, we propose a neural architecture search method based GCN (ASGCN) for the classification of PolSAR images. We construct a novel graph whose nodes combines both the physical features and spatial relations between pixels or samples to represent the image. Then we build a new searching space whose components are empirically selected from some graph neural networks for architecture search and develop the differentiable architecture search method to construction our ASGCN. Moreover, to address the training of large-scale images, we present a new weighted mini-batch algorithm to reduce the computing memory consumption and ensure the balance of sample distribution, and also analyze and compare with other similar training strategies. Experiments on several real-world PolSAR datasets show that our method has improved the overall accuracy as much as 3.76% than state-of-the-art methods.

Sign in / Sign up

Export Citation Format

Share Document