RaSEC : an intelligent framework for reliable and secure multilevel edge computing in industrial environments
- Usman, Muhammad, Jolfaei, Alireza, Jan, Mian
- Authors: Usman, Muhammad , Jolfaei, Alireza , Jan, Mian
- Date: 2020
- Type: Text , Journal article
- Relation: IEEE Transactions on Industry Applications Vol. 56, no. 4 (2020), p. 4543-4551
- Full Text:
- Reviewed:
- Description: Industrial applications generate big data with redundant information that is transmitted over heterogeneous networks. The transmission of big data with redundant information not only increases the overall end-to-end delay but also increases the computational load on servers which affects the performance of industrial applications. To address these challenges, we propose an intelligent framework named Reliable and Secure multi-level Edge Computing (RaSEC), which operates in three phases. In the first phase, level-one edge devices apply a lightweight aggregation technique on the generated data. This technique not only reduces the size of the generated data but also helps in preserving the privacy of data sources. In the second phase, a multistep process is used to register level-two edge devices (LTEDs) with high-level edge devices (HLEDs). Due to the registration process, only legitimate LTEDs can forward data to the HLEDs, and as a result, the computational load on HLEDs decreases. In the third phase, the HLEDs use a convolutional neural network to detect the presence of moving objects in the data forwarded by LTEDs. If a movement is detected, the data is uploaded to the cloud servers for further analysis; otherwise, the data is discarded to minimize the use of computational resources on cloud computing platforms. The proposed framework reduces the response time by forwarding useful information to the cloud servers and can be utilized by various industrial applications. Our theoretical and experimental results confirm the resiliency of our framework with respect to security and privacy threats. © 1972-2012 IEEE.
- Authors: Usman, Muhammad , Jolfaei, Alireza , Jan, Mian
- Date: 2020
- Type: Text , Journal article
- Relation: IEEE Transactions on Industry Applications Vol. 56, no. 4 (2020), p. 4543-4551
- Full Text:
- Reviewed:
- Description: Industrial applications generate big data with redundant information that is transmitted over heterogeneous networks. The transmission of big data with redundant information not only increases the overall end-to-end delay but also increases the computational load on servers which affects the performance of industrial applications. To address these challenges, we propose an intelligent framework named Reliable and Secure multi-level Edge Computing (RaSEC), which operates in three phases. In the first phase, level-one edge devices apply a lightweight aggregation technique on the generated data. This technique not only reduces the size of the generated data but also helps in preserving the privacy of data sources. In the second phase, a multistep process is used to register level-two edge devices (LTEDs) with high-level edge devices (HLEDs). Due to the registration process, only legitimate LTEDs can forward data to the HLEDs, and as a result, the computational load on HLEDs decreases. In the third phase, the HLEDs use a convolutional neural network to detect the presence of moving objects in the data forwarded by LTEDs. If a movement is detected, the data is uploaded to the cloud servers for further analysis; otherwise, the data is discarded to minimize the use of computational resources on cloud computing platforms. The proposed framework reduces the response time by forwarding useful information to the cloud servers and can be utilized by various industrial applications. Our theoretical and experimental results confirm the resiliency of our framework with respect to security and privacy threats. © 1972-2012 IEEE.
Real-time dissemination of emergency warning messages in 5G enabled selfish vehicular social networks
- Ullah, Noor, Kong, Xiangjie, Lin, Limei, Alrashoud, Mubarak, Tolba, Amr, Xia, Feng
- Authors: Ullah, Noor , Kong, Xiangjie , Lin, Limei , Alrashoud, Mubarak , Tolba, Amr , Xia, Feng
- Date: 2020
- Type: Text , Journal article
- Relation: Computer Networks Vol. 182, no. (2020), p.
- Full Text:
- Reviewed:
- Description: This paper addresses the issues of selfishness, limited network resources, and their adverse effects on real-time dissemination of Emergency Warning Messages (EWMs) in modern Autonomous Moving Platforms (AMPs) such as Vehicular Social Networks (VSNs). For this purpose, we propose a social intelligence based identification mechanism to differentiate between a selfish and a cooperative node in the network. Therefore, we devise a crowdsensing based mechanism to calculate a tie-strength value based on several social metrics. Moreover, we design a recursive evolutionary algorithm for each node's reputation calculation and update. Given that, then we estimate each node's state-transition probability to select a super-spreader for rapid dissemination. In order to ensure a seamless and reliable dissemination process, we incorporate 5G network structure instead of conventional short range communication which is used in most vehicular networks at present. Finally, we design a real-time dissemination algorithm for EWMs and evaluate its performance in terms of network parameters such as delivery-ratio, delay, hop-count, and message-overhead for varying values of vehicular density, speed, and selfish nodes’ density based on realistic vehicular mobility traces. In addition, we present a comparative analysis of the performance of the proposed scheme with state-of-the-art dissemination schemes in VSNs. © 2020 Elsevier B.V.
- Authors: Ullah, Noor , Kong, Xiangjie , Lin, Limei , Alrashoud, Mubarak , Tolba, Amr , Xia, Feng
- Date: 2020
- Type: Text , Journal article
- Relation: Computer Networks Vol. 182, no. (2020), p.
- Full Text:
- Reviewed:
- Description: This paper addresses the issues of selfishness, limited network resources, and their adverse effects on real-time dissemination of Emergency Warning Messages (EWMs) in modern Autonomous Moving Platforms (AMPs) such as Vehicular Social Networks (VSNs). For this purpose, we propose a social intelligence based identification mechanism to differentiate between a selfish and a cooperative node in the network. Therefore, we devise a crowdsensing based mechanism to calculate a tie-strength value based on several social metrics. Moreover, we design a recursive evolutionary algorithm for each node's reputation calculation and update. Given that, then we estimate each node's state-transition probability to select a super-spreader for rapid dissemination. In order to ensure a seamless and reliable dissemination process, we incorporate 5G network structure instead of conventional short range communication which is used in most vehicular networks at present. Finally, we design a real-time dissemination algorithm for EWMs and evaluate its performance in terms of network parameters such as delivery-ratio, delay, hop-count, and message-overhead for varying values of vehicular density, speed, and selfish nodes’ density based on realistic vehicular mobility traces. In addition, we present a comparative analysis of the performance of the proposed scheme with state-of-the-art dissemination schemes in VSNs. © 2020 Elsevier B.V.
Rectified softmax loss with all-sided cost sensitivity for age estimation
- Li, Daxiang, Ma, Xuan, Ren, Yaqiong, Teng, Shyh-Wei
- Authors: Li, Daxiang , Ma, Xuan , Ren, Yaqiong , Teng, Shyh-Wei
- Date: 2020
- Type: Text , Journal article
- Relation: IEEE Access Vol. 8, no. (2020), p. 32551-32563
- Full Text:
- Reviewed:
- Description: In Convolutional Neural Network (ConvNet) based age estimation algorithms, softmax loss is usually chosen as the loss function directly, and the problems of Cost Sensitivity (CS), such as class imbalance and misclassification cost difference between different classes, are not considered. Focus on these problems, this paper constructs a rectified softmax loss function with all-sided CS, and proposes a novel cost-sensitive ConvNet based age estimation algorithm. Firstly, a loss function is established for each age category to solve the imbalance of the number of training samples. Then, a cost matrix is defined to reflect the cost difference caused by misclassification between different classes, thus constructing a new cost-sensitive error function. Finally, the above methods are merged to construct a rectified softmax loss function for ConvNet model, and a corresponding Back Propagation (BP) training scheme is designed to enable ConvNet network to learn robust face representation for age estimation during the training phase. Simultaneously, the rectified softmax loss is theoretically proved that it satisfies the general conditions of the loss function used for classification. The effectiveness of the proposed method is verified by experiments on face image datasets of different races. © 2013 IEEE.
- Authors: Li, Daxiang , Ma, Xuan , Ren, Yaqiong , Teng, Shyh-Wei
- Date: 2020
- Type: Text , Journal article
- Relation: IEEE Access Vol. 8, no. (2020), p. 32551-32563
- Full Text:
- Reviewed:
- Description: In Convolutional Neural Network (ConvNet) based age estimation algorithms, softmax loss is usually chosen as the loss function directly, and the problems of Cost Sensitivity (CS), such as class imbalance and misclassification cost difference between different classes, are not considered. Focus on these problems, this paper constructs a rectified softmax loss function with all-sided CS, and proposes a novel cost-sensitive ConvNet based age estimation algorithm. Firstly, a loss function is established for each age category to solve the imbalance of the number of training samples. Then, a cost matrix is defined to reflect the cost difference caused by misclassification between different classes, thus constructing a new cost-sensitive error function. Finally, the above methods are merged to construct a rectified softmax loss function for ConvNet model, and a corresponding Back Propagation (BP) training scheme is designed to enable ConvNet network to learn robust face representation for age estimation during the training phase. Simultaneously, the rectified softmax loss is theoretically proved that it satisfies the general conditions of the loss function used for classification. The effectiveness of the proposed method is verified by experiments on face image datasets of different races. © 2013 IEEE.
Reduced switch multilevel inverter topologies for renewable energy sources
- Sarebanzadeh, Maryam, Hosseinzadeh, Mohammad, Garcia, Cristian, Babaei, Ebrahim, Islam, Syed, Rodriguez, Jose
- Authors: Sarebanzadeh, Maryam , Hosseinzadeh, Mohammad , Garcia, Cristian , Babaei, Ebrahim , Islam, Syed , Rodriguez, Jose
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Access Vol. 9, no. (2021), p. 120580-120595
- Full Text:
- Reviewed:
- Description: This article proposes two generalized multilevel inverter configurations that reduce the number of switching devices, isolated DC sources, and total standing voltage on power switches, making them suitable for renewable energy sources. The main topology is a multilevel inverter that handles two isolated DC sources with ten power switches to create 25 voltage levels. Based on the main proposed topology, two generalized multilevel inverters are introduced to provide flexibility in the design and to minimize the number of elements. The optimal topologies for both extensive multilevel inverters are derived from different design objectives such as minimizing the number of elements (gate drivers, DC sources), achieving a large number of levels, and minimizing the total standing voltage. The main advantages of the proposed topologies are a reduced number of elements compared to those required by other existing multilevel inverter topologies. The power loss analysis and standalone PV application of the proposed topologies are discussed. Experimental results are presented for the proposed topology to demonstrate its correct operation. © 2013 IEEE.
- Authors: Sarebanzadeh, Maryam , Hosseinzadeh, Mohammad , Garcia, Cristian , Babaei, Ebrahim , Islam, Syed , Rodriguez, Jose
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Access Vol. 9, no. (2021), p. 120580-120595
- Full Text:
- Reviewed:
- Description: This article proposes two generalized multilevel inverter configurations that reduce the number of switching devices, isolated DC sources, and total standing voltage on power switches, making them suitable for renewable energy sources. The main topology is a multilevel inverter that handles two isolated DC sources with ten power switches to create 25 voltage levels. Based on the main proposed topology, two generalized multilevel inverters are introduced to provide flexibility in the design and to minimize the number of elements. The optimal topologies for both extensive multilevel inverters are derived from different design objectives such as minimizing the number of elements (gate drivers, DC sources), achieving a large number of levels, and minimizing the total standing voltage. The main advantages of the proposed topologies are a reduced number of elements compared to those required by other existing multilevel inverter topologies. The power loss analysis and standalone PV application of the proposed topologies are discussed. Experimental results are presented for the proposed topology to demonstrate its correct operation. © 2013 IEEE.
Reusing artifact-centric business process models : a behavioral consistent specialization approach
- Yongchareon, Sira, Liu, Chengfei, Zhao, Xiaohui
- Authors: Yongchareon, Sira , Liu, Chengfei , Zhao, Xiaohui
- Date: 2020
- Type: Text , Journal article
- Relation: Computing Vol. 102, no. 8 (2020), p. 1843-1879
- Full Text:
- Reviewed:
- Description: Process reuse is one of the important research areas that address efficiency issues in business process modeling. Similar to software reuse, business processes should be able to be componentized and specialized in order to enable flexible process expansion and customization. Current activity/control-flow centric workflow modeling approaches face difficulty in supporting highly flexible process reuse, limited by their procedural nature. In comparison, the emerging artifact-centric workflow modeling approach well fits into these reuse requirements. Beyond the classic class level reuse in existing object-oriented approaches, process reuse faces the challenge of handling synchronization dependencies among artifact lifecycles as parts of a business process. In this article, we propose a theoretical framework for business process specialization that comprises an artifact-centric business process model, a set of methods to design and construct a specialized business process model from a base model, and a set of behavioral consistency criteria to help check the consistency between the two process models. © 2020, Springer-Verlag GmbH Austria, part of Springer Nature.
- Authors: Yongchareon, Sira , Liu, Chengfei , Zhao, Xiaohui
- Date: 2020
- Type: Text , Journal article
- Relation: Computing Vol. 102, no. 8 (2020), p. 1843-1879
- Full Text:
- Reviewed:
- Description: Process reuse is one of the important research areas that address efficiency issues in business process modeling. Similar to software reuse, business processes should be able to be componentized and specialized in order to enable flexible process expansion and customization. Current activity/control-flow centric workflow modeling approaches face difficulty in supporting highly flexible process reuse, limited by their procedural nature. In comparison, the emerging artifact-centric workflow modeling approach well fits into these reuse requirements. Beyond the classic class level reuse in existing object-oriented approaches, process reuse faces the challenge of handling synchronization dependencies among artifact lifecycles as parts of a business process. In this article, we propose a theoretical framework for business process specialization that comprises an artifact-centric business process model, a set of methods to design and construct a specialized business process model from a base model, and a set of behavioral consistency criteria to help check the consistency between the two process models. © 2020, Springer-Verlag GmbH Austria, part of Springer Nature.
Robust image classification using a low-pass activation function and DCT augmentation
- Hossain, Md Tahmid, Teng, Shyh, Sohel, Ferdous, Lu, Guojun
- Authors: Hossain, Md Tahmid , Teng, Shyh , Sohel, Ferdous , Lu, Guojun
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Access Vol. 9, no. (2021), p. 86460-86474
- Full Text:
- Reviewed:
- Description: Convolutional Neural Network's (CNN's) performance disparity on clean and corrupted datasets has recently come under scrutiny. In this work, we analyse common corruptions in the frequency domain, i.e., High Frequency corruptions (HFc, e.g., noise) and Low Frequency corruptions (LFc, e.g., blur). Although a simple solution to HFc is low-pass filtering, ReLU - a widely used Activation Function (AF), does not have any filtering mechanism. In this work, we instill low-pass filtering into the AF (LP-ReLU) to improve robustness against HFc. To deal with LFc, we complement LP-ReLU with Discrete Cosine Transform based augmentation. LP-ReLU, coupled with DCT augmentation, enables a deep network to tackle the entire spectrum of corruption. We use CIFAR-10-C and Tiny ImageNet-C for evaluation and demonstrate improvements of 5% and 7.3% in accuracy respectively, compared to the State-Of-The-Art (SOTA). We further evaluate our method's stability on a variety of perturbations in CIFAR-10-P and Tiny ImageNet-P, achieving new SOTA in these experiments as well. To further strengthen our understanding regarding CNN's lack of robustness, a decision space visualisation process is proposed and presented in this work. © 2013 IEEE.
- Authors: Hossain, Md Tahmid , Teng, Shyh , Sohel, Ferdous , Lu, Guojun
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Access Vol. 9, no. (2021), p. 86460-86474
- Full Text:
- Reviewed:
- Description: Convolutional Neural Network's (CNN's) performance disparity on clean and corrupted datasets has recently come under scrutiny. In this work, we analyse common corruptions in the frequency domain, i.e., High Frequency corruptions (HFc, e.g., noise) and Low Frequency corruptions (LFc, e.g., blur). Although a simple solution to HFc is low-pass filtering, ReLU - a widely used Activation Function (AF), does not have any filtering mechanism. In this work, we instill low-pass filtering into the AF (LP-ReLU) to improve robustness against HFc. To deal with LFc, we complement LP-ReLU with Discrete Cosine Transform based augmentation. LP-ReLU, coupled with DCT augmentation, enables a deep network to tackle the entire spectrum of corruption. We use CIFAR-10-C and Tiny ImageNet-C for evaluation and demonstrate improvements of 5% and 7.3% in accuracy respectively, compared to the State-Of-The-Art (SOTA). We further evaluate our method's stability on a variety of perturbations in CIFAR-10-P and Tiny ImageNet-P, achieving new SOTA in these experiments as well. To further strengthen our understanding regarding CNN's lack of robustness, a decision space visualisation process is proposed and presented in this work. © 2013 IEEE.
Robust malware defense in industrial IoT applications using machine learning with selective adversarial samples
- Khoda, Mahbub, Imam, Tasadduq, Kamruzzaman, Joarder, Gondal, Iqbal, Rahman, Ashfaqur
- Authors: Khoda, Mahbub , Imam, Tasadduq , Kamruzzaman, Joarder , Gondal, Iqbal , Rahman, Ashfaqur
- Date: 2019
- Type: Text , Journal article
- Relation: IEEE Transactions on Industry Applications Vol.56, no 4. (2020), p. 4415-4424
- Full Text:
- Reviewed:
- Description: Industrial Internet of Things (IIoT) deploys edge devices to act as intermediaries between sensors and actuators and application servers or cloud services. Machine learning models have been widely used to thwart malware attacks in such edge devices. However, these models are vulnerable to adversarial attacks where attackers craft adversarial samples by introducing small perturbations to malware samples to fool a classifier to misclassify them as benign applications. Literature on deep learning networks proposes adversarial retraining as a defense mechanism where adversarial samples are combined with legitimate samples to retrain the classifier. However, existing works select such adversarial samples in a random fashion which degrades the classifier's performance. This work proposes two novel approaches for selecting adversarial samples to retrain a classifier. One, based on the distance from malware cluster center, and the other, based on a probability measure derived from a kernel based learning (KBL). Our experiments show that both of our sample selection methods outperform the random selection method and the KBL selection method improves detection accuracy by 6%. Also, while existing works focus on deep neural networks with respect to adversarial retraining, we additionally assess the impact of such adversarial samples on other classifiers and our proposed selective adversarial retraining approaches show similar performance improvement for these classifiers as well. The outcomes from the study can assist in designing robust security systems for IIoT applications.
- Authors: Khoda, Mahbub , Imam, Tasadduq , Kamruzzaman, Joarder , Gondal, Iqbal , Rahman, Ashfaqur
- Date: 2019
- Type: Text , Journal article
- Relation: IEEE Transactions on Industry Applications Vol.56, no 4. (2020), p. 4415-4424
- Full Text:
- Reviewed:
- Description: Industrial Internet of Things (IIoT) deploys edge devices to act as intermediaries between sensors and actuators and application servers or cloud services. Machine learning models have been widely used to thwart malware attacks in such edge devices. However, these models are vulnerable to adversarial attacks where attackers craft adversarial samples by introducing small perturbations to malware samples to fool a classifier to misclassify them as benign applications. Literature on deep learning networks proposes adversarial retraining as a defense mechanism where adversarial samples are combined with legitimate samples to retrain the classifier. However, existing works select such adversarial samples in a random fashion which degrades the classifier's performance. This work proposes two novel approaches for selecting adversarial samples to retrain a classifier. One, based on the distance from malware cluster center, and the other, based on a probability measure derived from a kernel based learning (KBL). Our experiments show that both of our sample selection methods outperform the random selection method and the KBL selection method improves detection accuracy by 6%. Also, while existing works focus on deep neural networks with respect to adversarial retraining, we additionally assess the impact of such adversarial samples on other classifiers and our proposed selective adversarial retraining approaches show similar performance improvement for these classifiers as well. The outcomes from the study can assist in designing robust security systems for IIoT applications.
Rock-burst occurrence prediction based on optimized naïve bayes models
- Ke, Bo, Khandelwal, Manoj, Asteris, Panagiotis, Skentou, Athanasia, Mamou, Anna, Armaghani, Danial
- Authors: Ke, Bo , Khandelwal, Manoj , Asteris, Panagiotis , Skentou, Athanasia , Mamou, Anna , Armaghani, Danial
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Access Vol. 9, no. (2021), p. 91347-91360
- Full Text:
- Reviewed:
- Description: Rock-burst is a common failure in hard rock related projects in civil and mining construction and therefore, proper classification and prediction of this phenomenon is of interest. This research presents the development of optimized naïve Bayes models, in predicting rock-burst failures in underground projects. The naïve Bayes models were optimized using four weight optimization techniques including forward, backward, particle swarm optimization, and evolutionary. An evolutionary random forest model was developed to identify the most significant input parameters. The maximum tangential stress, elastic energy index, and uniaxial tensile stress were then selected by the feature selection technique (i.e., evolutionary random forest) to develop the optimized naïve Bayes models. The performance of the models was assessed using various criteria as well as a simple ranking system. The results of this research showed that particle swarm optimization was the most effective technique in improving the accuracy of the naïve Bayes model for rock-burst prediction (cumulative ranking = 21), while the backward technique was the worst weight optimization technique (cumulative ranking = 11). All the optimized naïve Bayes models identified the maximum tangential stress as the most significant parameter in predicting rock-burst failures. The results of this research demonstrate that particle swarm optimization technique may improve the accuracy of naïve Bayes algorithms in predicting rock-burst occurrence. © 2013 IEEE.
- Authors: Ke, Bo , Khandelwal, Manoj , Asteris, Panagiotis , Skentou, Athanasia , Mamou, Anna , Armaghani, Danial
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Access Vol. 9, no. (2021), p. 91347-91360
- Full Text:
- Reviewed:
- Description: Rock-burst is a common failure in hard rock related projects in civil and mining construction and therefore, proper classification and prediction of this phenomenon is of interest. This research presents the development of optimized naïve Bayes models, in predicting rock-burst failures in underground projects. The naïve Bayes models were optimized using four weight optimization techniques including forward, backward, particle swarm optimization, and evolutionary. An evolutionary random forest model was developed to identify the most significant input parameters. The maximum tangential stress, elastic energy index, and uniaxial tensile stress were then selected by the feature selection technique (i.e., evolutionary random forest) to develop the optimized naïve Bayes models. The performance of the models was assessed using various criteria as well as a simple ranking system. The results of this research showed that particle swarm optimization was the most effective technique in improving the accuracy of the naïve Bayes model for rock-burst prediction (cumulative ranking = 21), while the backward technique was the worst weight optimization technique (cumulative ranking = 11). All the optimized naïve Bayes models identified the maximum tangential stress as the most significant parameter in predicting rock-burst failures. The results of this research demonstrate that particle swarm optimization technique may improve the accuracy of naïve Bayes algorithms in predicting rock-burst occurrence. © 2013 IEEE.
Shifu2 : a network representation learning based model for advisor-advisee relationship mining
- Liu, Jiaying, Xia, Feng, Wang, Lei, Xu, Bo, Kong, Xiangjie
- Authors: Liu, Jiaying , Xia, Feng , Wang, Lei , Xu, Bo , Kong, Xiangjie
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Transactions on Knowledge and Data Engineering Vol. 33, no. 4 (2021), p. 1763-1777
- Full Text:
- Reviewed:
- Description: The advisor-advisee relationship represents direct knowledge heritage, and such relationship may not be readily available from academic libraries and search engines. This work aims to discover advisor-advisee relationships hidden behind scientific collaboration networks. For this purpose, we propose a novel model based on Network Representation Learning (NRL), namely Shifu2, which takes the collaboration network as input and the identified advisor-advisee relationship as output. In contrast to existing NRL models, Shifu2 considers not only the network structure but also the semantic information of nodes and edges. Shifu2 encodes nodes and edges into low-dimensional vectors respectively, both of which are then utilized to identify advisor-advisee relationships. Experimental results illustrate improved stability and effectiveness of the proposed model over state-of-the-art methods. In addition, we generate a large-scale academic genealogy dataset by taking advantage of Shifu2. © 1989-2012 IEEE.
- Authors: Liu, Jiaying , Xia, Feng , Wang, Lei , Xu, Bo , Kong, Xiangjie
- Date: 2021
- Type: Text , Journal article
- Relation: IEEE Transactions on Knowledge and Data Engineering Vol. 33, no. 4 (2021), p. 1763-1777
- Full Text:
- Reviewed:
- Description: The advisor-advisee relationship represents direct knowledge heritage, and such relationship may not be readily available from academic libraries and search engines. This work aims to discover advisor-advisee relationships hidden behind scientific collaboration networks. For this purpose, we propose a novel model based on Network Representation Learning (NRL), namely Shifu2, which takes the collaboration network as input and the identified advisor-advisee relationship as output. In contrast to existing NRL models, Shifu2 considers not only the network structure but also the semantic information of nodes and edges. Shifu2 encodes nodes and edges into low-dimensional vectors respectively, both of which are then utilized to identify advisor-advisee relationships. Experimental results illustrate improved stability and effectiveness of the proposed model over state-of-the-art methods. In addition, we generate a large-scale academic genealogy dataset by taking advantage of Shifu2. © 1989-2012 IEEE.
Simplifying and improving ant-based clustering
- Tan, Swee, Ting, Kaiming, Teng, Shyh
- Authors: Tan, Swee , Ting, Kaiming , Teng, Shyh
- Date: 2011
- Type: Text , Conference paper
- Relation: 11th International Conference on Computational Science, ICCS 2011; Singapore, Singapore; 1st-3rd June 2011, published in Procedia Computer Science Vol. 4, p. 46-55
- Full Text:
- Reviewed:
- Description: Ant-based clustering (ABC) is a data clustering approach inspired from cemetery formation activities observed in real ant colonies. Building upon the premise of collective intelligence, such an approach uses multiple ant-like agents and a mixture of heuristics, in order to create systems that are capable of clustering real-world data. Many recently proposed ABC systems have shown competitive results, but these systems are geared towards adding new heuristics, resulting in increasingly complex systems that are harder to understand and improve. In contrast to this direction, we demonstrate that a state-of-the-art ABC system can be systematically evaluated and then simplified. The streamlined model, which we call SABC, differs fundamentally from traditional ABC systems as it does not use the ant-colony and several key components. Yet, our empirical study shows that SABC performs more effectively and effciently than the state-of-the-art ABC system.
- Authors: Tan, Swee , Ting, Kaiming , Teng, Shyh
- Date: 2011
- Type: Text , Conference paper
- Relation: 11th International Conference on Computational Science, ICCS 2011; Singapore, Singapore; 1st-3rd June 2011, published in Procedia Computer Science Vol. 4, p. 46-55
- Full Text:
- Reviewed:
- Description: Ant-based clustering (ABC) is a data clustering approach inspired from cemetery formation activities observed in real ant colonies. Building upon the premise of collective intelligence, such an approach uses multiple ant-like agents and a mixture of heuristics, in order to create systems that are capable of clustering real-world data. Many recently proposed ABC systems have shown competitive results, but these systems are geared towards adding new heuristics, resulting in increasingly complex systems that are harder to understand and improve. In contrast to this direction, we demonstrate that a state-of-the-art ABC system can be systematically evaluated and then simplified. The streamlined model, which we call SABC, differs fundamentally from traditional ABC systems as it does not use the ant-colony and several key components. Yet, our empirical study shows that SABC performs more effectively and effciently than the state-of-the-art ABC system.
Softmax exploration strategies for multiobjective reinforcement learning
- Vamplew, Peter, Dazeley, Richard, Foale, Cameron
- Authors: Vamplew, Peter , Dazeley, Richard , Foale, Cameron
- Date: 2017
- Type: Text , Journal article
- Relation: Neurocomputing Vol. 263, no. (2017), p. 74-86
- Full Text:
- Reviewed:
- Description: Despite growing interest over recent years in applying reinforcement learning to multiobjective problems, there has been little research into the applicability and effectiveness of exploration strategies within the multiobjective context. This work considers several widely-used approaches to exploration from the single-objective reinforcement learning literature, and examines their incorporation into multiobjective Q-learning. In particular this paper proposes two novel approaches which extend the softmax operator to work with vector-valued rewards. The performance of these exploration strategies is evaluated across a set of benchmark environments. Issues arising from the multiobjective formulation of these benchmarks which impact on the performance of the exploration strategies are identified. It is shown that of the techniques considered, the combination of the novel softmax–epsilon exploration with optimistic initialisation provides the most effective trade-off between exploration and exploitation.
- Authors: Vamplew, Peter , Dazeley, Richard , Foale, Cameron
- Date: 2017
- Type: Text , Journal article
- Relation: Neurocomputing Vol. 263, no. (2017), p. 74-86
- Full Text:
- Reviewed:
- Description: Despite growing interest over recent years in applying reinforcement learning to multiobjective problems, there has been little research into the applicability and effectiveness of exploration strategies within the multiobjective context. This work considers several widely-used approaches to exploration from the single-objective reinforcement learning literature, and examines their incorporation into multiobjective Q-learning. In particular this paper proposes two novel approaches which extend the softmax operator to work with vector-valued rewards. The performance of these exploration strategies is evaluated across a set of benchmark environments. Issues arising from the multiobjective formulation of these benchmarks which impact on the performance of the exploration strategies are identified. It is shown that of the techniques considered, the combination of the novel softmax–epsilon exploration with optimistic initialisation provides the most effective trade-off between exploration and exploitation.
Steering approaches to Pareto-optimal multiobjective reinforcement learning
- Vamplew, Peter, Issabekov, Rustam, Dazeley, Richard, Foale, Cameron, Berry, Adam, Moore, Tim, Creighton, Douglas
- Authors: Vamplew, Peter , Issabekov, Rustam , Dazeley, Richard , Foale, Cameron , Berry, Adam , Moore, Tim , Creighton, Douglas
- Date: 2017
- Type: Text , Journal article
- Relation: Neurocomputing Vol. 263, no. (2017), p. 26-38
- Full Text:
- Reviewed:
- Description: For reinforcement learning tasks with multiple objectives, it may be advantageous to learn stochastic or non-stationary policies. This paper investigates two novel algorithms for learning non-stationary policies which produce Pareto-optimal behaviour (w-steering and Q-steering), by extending prior work based on the concept of geometric steering. Empirical results demonstrate that both new algorithms offer substantial performance improvements over stationary deterministic policies, while Q-steering significantly outperforms w-steering when the agent has no information about recurrent states within the environment. It is further demonstrated that Q-steering can be used interactively by providing a human decision-maker with a visualisation of the Pareto front and allowing them to adjust the agent’s target point during learning. To demonstrate broader applicability, the use of Q-steering in combination with function approximation is also illustrated on a task involving control of local battery storage for a residential solar power system.
- Authors: Vamplew, Peter , Issabekov, Rustam , Dazeley, Richard , Foale, Cameron , Berry, Adam , Moore, Tim , Creighton, Douglas
- Date: 2017
- Type: Text , Journal article
- Relation: Neurocomputing Vol. 263, no. (2017), p. 26-38
- Full Text:
- Reviewed:
- Description: For reinforcement learning tasks with multiple objectives, it may be advantageous to learn stochastic or non-stationary policies. This paper investigates two novel algorithms for learning non-stationary policies which produce Pareto-optimal behaviour (w-steering and Q-steering), by extending prior work based on the concept of geometric steering. Empirical results demonstrate that both new algorithms offer substantial performance improvements over stationary deterministic policies, while Q-steering significantly outperforms w-steering when the agent has no information about recurrent states within the environment. It is further demonstrated that Q-steering can be used interactively by providing a human decision-maker with a visualisation of the Pareto front and allowing them to adjust the agent’s target point during learning. To demonstrate broader applicability, the use of Q-steering in combination with function approximation is also illustrated on a task involving control of local battery storage for a residential solar power system.
The diagnostic validity and reliability of an internet-based clinical assessment program for mental disorders
- Nguyen, David, Klein, Britt, Meyer, Denny, Austin, David, Abbott, Jo-Anne
- Authors: Nguyen, David , Klein, Britt , Meyer, Denny , Austin, David , Abbott, Jo-Anne
- Date: 2015
- Type: Text , Journal article
- Relation: Journal of Medical Internet Research Vol. 17, no. 9 (2015), p.
- Full Text:
- Reviewed:
- Description: Background: Internet-based assessment has the potential to assist with the diagnosis of mental health disorders and overcome the barriers associated with traditional services (eg, cost, stigma, distance). Further to existing online screening programs available, there is an opportunity to deliver more comprehensive and accurate diagnostic tools to supplement the assessment and treatment of mental health disorders. Objective: The aim was to evaluate the diagnostic criterion validity and test-retest reliability of the electronic Psychological Assessment System (e-PASS), an online, self-report, multidisorder, clinical assessment and referral system. Methods: Participants were 616 adults residing in Australia, recruited online, and representing prospective e-PASS users. Following e-PASS completion, 158 participants underwent a telephone-administered structured clinical interview and 39 participants repeated the e-PASS within 25 days of initial completion. Results: With structured clinical interview results serving as the gold standard, diagnostic agreement with the e-PASS varied considerably from fair (eg, generalized anxiety disorder:kappa=.37) to strong (eg, panic disorder:kappa=.62). Although the e-PASS' sensitivity also varied (0.43-0.86) the specificity was generally high (0.68-1.00). The e-PASS sensitivity generally improved when reducing the e-PASS threshold to a subclinical result. Test-retest reliability ranged from moderate (eg, specific phobia:kappa=.54) to substantial (eg, bulimia nervosa:kappa=.87). Conclusions: The e-PASS produces reliable diagnostic results and performs generally well in excluding mental disorders, although at the expense of sensitivity. For screening purposes, the e-PASS subclinical result generally appears better than a clinical result as a diagnostic indicator. Further development and evaluation is needed to support the use of online diagnostic assessment programs for mental disorders.
- Authors: Nguyen, David , Klein, Britt , Meyer, Denny , Austin, David , Abbott, Jo-Anne
- Date: 2015
- Type: Text , Journal article
- Relation: Journal of Medical Internet Research Vol. 17, no. 9 (2015), p.
- Full Text:
- Reviewed:
- Description: Background: Internet-based assessment has the potential to assist with the diagnosis of mental health disorders and overcome the barriers associated with traditional services (eg, cost, stigma, distance). Further to existing online screening programs available, there is an opportunity to deliver more comprehensive and accurate diagnostic tools to supplement the assessment and treatment of mental health disorders. Objective: The aim was to evaluate the diagnostic criterion validity and test-retest reliability of the electronic Psychological Assessment System (e-PASS), an online, self-report, multidisorder, clinical assessment and referral system. Methods: Participants were 616 adults residing in Australia, recruited online, and representing prospective e-PASS users. Following e-PASS completion, 158 participants underwent a telephone-administered structured clinical interview and 39 participants repeated the e-PASS within 25 days of initial completion. Results: With structured clinical interview results serving as the gold standard, diagnostic agreement with the e-PASS varied considerably from fair (eg, generalized anxiety disorder:kappa=.37) to strong (eg, panic disorder:kappa=.62). Although the e-PASS' sensitivity also varied (0.43-0.86) the specificity was generally high (0.68-1.00). The e-PASS sensitivity generally improved when reducing the e-PASS threshold to a subclinical result. Test-retest reliability ranged from moderate (eg, specific phobia:kappa=.54) to substantial (eg, bulimia nervosa:kappa=.87). Conclusions: The e-PASS produces reliable diagnostic results and performs generally well in excluding mental disorders, although at the expense of sensitivity. For screening purposes, the e-PASS subclinical result generally appears better than a clinical result as a diagnostic indicator. Further development and evaluation is needed to support the use of online diagnostic assessment programs for mental disorders.
The impact of handwriting difficulties on compositional quality in children with developmental coordination disorder
- Prunty, Mellissa, Barnett, Anna, Wilmut, Kate, Plumb, Mandy
- Authors: Prunty, Mellissa , Barnett, Anna , Wilmut, Kate , Plumb, Mandy
- Date: 2016
- Type: Text , Journal article
- Relation: British Journal of Occupational Therapy Vol. 79, no. 10 (2016), p. 591-597
- Full Text:
- Reviewed:
- Description: Introduction There is substantial evidence to support the relationship between transcription skills (handwriting and spelling) and compositional quality. For children with developmental coordination disorder, handwriting can be particularly challenging. While recent research has aimed to investigate their handwriting difficulties in more detail, the impact of transcription on their compositional quality has not previously been examined. The aim of this exploratory study was to examine compositional quality in children with developmental coordination disorder and to ascertain whether their transcription skills influence writing quality. Method Twenty-eight children with developmental coordination disorder participated in the study, with 28 typically developing age and gender matched controls. The children completed the free-writing' task from the detailed assessment of speed of handwriting tool, which was evaluated for compositional quality using the Wechsler objective language dimensions. Results The children with developmental coordination disorder performed significantly below their typically developing peers on five of the six Wechsler objective language dimensions items. They also had a higher percentage of misspelled words. Regression analyses indicated that the number of words produced per minute and the percentage of misspelled words explained 55% of the variance for compositional quality. Conclusion The handwriting difficulties so commonly reported in children with developmental coordination disorder have wider repercussions for the quality of written composition.
- Authors: Prunty, Mellissa , Barnett, Anna , Wilmut, Kate , Plumb, Mandy
- Date: 2016
- Type: Text , Journal article
- Relation: British Journal of Occupational Therapy Vol. 79, no. 10 (2016), p. 591-597
- Full Text:
- Reviewed:
- Description: Introduction There is substantial evidence to support the relationship between transcription skills (handwriting and spelling) and compositional quality. For children with developmental coordination disorder, handwriting can be particularly challenging. While recent research has aimed to investigate their handwriting difficulties in more detail, the impact of transcription on their compositional quality has not previously been examined. The aim of this exploratory study was to examine compositional quality in children with developmental coordination disorder and to ascertain whether their transcription skills influence writing quality. Method Twenty-eight children with developmental coordination disorder participated in the study, with 28 typically developing age and gender matched controls. The children completed the free-writing' task from the detailed assessment of speed of handwriting tool, which was evaluated for compositional quality using the Wechsler objective language dimensions. Results The children with developmental coordination disorder performed significantly below their typically developing peers on five of the six Wechsler objective language dimensions items. They also had a higher percentage of misspelled words. Regression analyses indicated that the number of words produced per minute and the percentage of misspelled words explained 55% of the variance for compositional quality. Conclusion The handwriting difficulties so commonly reported in children with developmental coordination disorder have wider repercussions for the quality of written composition.
Therapist-Assisted, Internet-Based Treatment for Panic Disorder: Can General Practitioners achieve comparable patient outcomes to Psychologists?
- Shandley, Kerrie, Austin, David, Klein, Britt, Pier, Ciaran, Schattner, Peter, Pierce, David, Wade, Victoria
- Authors: Shandley, Kerrie , Austin, David , Klein, Britt , Pier, Ciaran , Schattner, Peter , Pierce, David , Wade, Victoria
- Date: 2008
- Type: Text , Journal article
- Relation: Journal of Medical Internet Research Vol. 10, no. 2 (2008), p. 1-15
- Full Text:
- Reviewed:
- Description: Background: Mental illness is an escalating concern worldwide. The management of disorders such as anxiety and depression largely falls to family doctors or general practitioners (GPs). However, GPs are often too time constrained and may lack the necessary training to adequately manage the needs of such patients. Evidence-based Internet interventions represent a potentially valuable resource to reduce the burden of care and the cost of managing mental health disorders within primary care settings and, at the same time, improve patient outcomes. Objective: The present study sought to extend the efficacy of a therapist-assisted Internet treatment program for panic disorder, Panic Online, by determining whether comparable outcomes could be achieved and maintained when Panic Online was supported by either GPs or psychologists. Methods: Via a natural groups design, 96 people with a primary diagnosis of panic disorder (with or without agoraphobia) completed the Panic Online program over 12 weeks with the therapeutic assistance of their GP (n = 53), who had received specialist training in cognitive behavioral therapy, or a clinical psychologist (n = 43). Participants completed a clinical diagnostic telephone interview, conducted by a psychologist, and a set of online questionnaires to assess panic-related symptoms at three time periods (pretreatment, posttreatment, and 6 month follow-up). Results: Both treatments led to clinically significant improvements on measures of panic and panic-related symptomatology from pretreatment to posttreatment. Both groups were shown to significantly improve over time. Improvements for both groups were maintained at follow-up; however, the groups did differ significantly on two quality of life domains: physical (F1,82 = 9.13, P = .00) and environmental (F1,82 = 4.41, P = .04). The attrition rate was significantly higher among those being treated by their GP (χ 2 1 = 4.40, P = .02, N = 96). Conclusions: This study provides evidence that Internet-based interventions are an effective adjunct to existing mental health care systems. Consequently, this may facilitate and enhance the delivery of evidence-based mental health treatments to increasingly large segments of the population via primary care systems and through suitably trained health professionals.
- Description: Background: Mental illness is an escalating concern worldwide. The management of disorders such as anxiety and depression largely falls to family doctors or general practitioners (GPs). However, GPs are often too time constrained and may lack the necessary training to adequately manage the needs of such patients. Evidence-based Internet interventions represent a potentially valuable resource to reduce the burden of care and the cost of managing mental health disorders within primary care settings and, at the same time, improve patient outcomes. Objective: The present study sought to extend the efficacy of a therapist-assisted Internet treatment program for panic disorder, Panic Online, by determining whether comparable outcomes could be achieved and maintained when Panic Online was supported by either GPs or psychologists. Methods: Via a natural groups design, 96 people with a primary diagnosis of panic disorder (with or without agoraphobia) completed the Panic Online program over 12 weeks with the therapeutic assistance of their GP (n = 53), who had received specialist training in cognitive behavioral therapy, or a clinical psychologist (n = 43). Participants completed a clinical diagnostic telephone interview, conducted by a psychologist, and a set of online questionnaires to assess panic-related symptoms at three time periods (pretreatment, posttreatment, and 6 month follow-up). Results: Both treatments led to clinically significant improvements on measures of panic and panic-related symptomatology from pretreatment to posttreatment. Both groups were shown to significantly improve over time. Improvements for both groups were maintained at follow-up; however, the groups did differ significantly on two quality of life domains: physical (F1,82 = 9.13, P = .00) and environmental (F1,82 = 4.41, P = .04). The attrition rate was significantly higher among those being treated by their GP (
- Authors: Shandley, Kerrie , Austin, David , Klein, Britt , Pier, Ciaran , Schattner, Peter , Pierce, David , Wade, Victoria
- Date: 2008
- Type: Text , Journal article
- Relation: Journal of Medical Internet Research Vol. 10, no. 2 (2008), p. 1-15
- Full Text:
- Reviewed:
- Description: Background: Mental illness is an escalating concern worldwide. The management of disorders such as anxiety and depression largely falls to family doctors or general practitioners (GPs). However, GPs are often too time constrained and may lack the necessary training to adequately manage the needs of such patients. Evidence-based Internet interventions represent a potentially valuable resource to reduce the burden of care and the cost of managing mental health disorders within primary care settings and, at the same time, improve patient outcomes. Objective: The present study sought to extend the efficacy of a therapist-assisted Internet treatment program for panic disorder, Panic Online, by determining whether comparable outcomes could be achieved and maintained when Panic Online was supported by either GPs or psychologists. Methods: Via a natural groups design, 96 people with a primary diagnosis of panic disorder (with or without agoraphobia) completed the Panic Online program over 12 weeks with the therapeutic assistance of their GP (n = 53), who had received specialist training in cognitive behavioral therapy, or a clinical psychologist (n = 43). Participants completed a clinical diagnostic telephone interview, conducted by a psychologist, and a set of online questionnaires to assess panic-related symptoms at three time periods (pretreatment, posttreatment, and 6 month follow-up). Results: Both treatments led to clinically significant improvements on measures of panic and panic-related symptomatology from pretreatment to posttreatment. Both groups were shown to significantly improve over time. Improvements for both groups were maintained at follow-up; however, the groups did differ significantly on two quality of life domains: physical (F1,82 = 9.13, P = .00) and environmental (F1,82 = 4.41, P = .04). The attrition rate was significantly higher among those being treated by their GP (χ 2 1 = 4.40, P = .02, N = 96). Conclusions: This study provides evidence that Internet-based interventions are an effective adjunct to existing mental health care systems. Consequently, this may facilitate and enhance the delivery of evidence-based mental health treatments to increasingly large segments of the population via primary care systems and through suitably trained health professionals.
- Description: Background: Mental illness is an escalating concern worldwide. The management of disorders such as anxiety and depression largely falls to family doctors or general practitioners (GPs). However, GPs are often too time constrained and may lack the necessary training to adequately manage the needs of such patients. Evidence-based Internet interventions represent a potentially valuable resource to reduce the burden of care and the cost of managing mental health disorders within primary care settings and, at the same time, improve patient outcomes. Objective: The present study sought to extend the efficacy of a therapist-assisted Internet treatment program for panic disorder, Panic Online, by determining whether comparable outcomes could be achieved and maintained when Panic Online was supported by either GPs or psychologists. Methods: Via a natural groups design, 96 people with a primary diagnosis of panic disorder (with or without agoraphobia) completed the Panic Online program over 12 weeks with the therapeutic assistance of their GP (n = 53), who had received specialist training in cognitive behavioral therapy, or a clinical psychologist (n = 43). Participants completed a clinical diagnostic telephone interview, conducted by a psychologist, and a set of online questionnaires to assess panic-related symptoms at three time periods (pretreatment, posttreatment, and 6 month follow-up). Results: Both treatments led to clinically significant improvements on measures of panic and panic-related symptomatology from pretreatment to posttreatment. Both groups were shown to significantly improve over time. Improvements for both groups were maintained at follow-up; however, the groups did differ significantly on two quality of life domains: physical (F1,82 = 9.13, P = .00) and environmental (F1,82 = 4.41, P = .04). The attrition rate was significantly higher among those being treated by their GP (
TOSNet : a topic-based optimal subnetwork identification in academic networks
- Bedru, Hayat, Zhao, Wenhong, Alrashoud, Mubarak, Tolba, Amr, Guo, He, Xia, Feng
- Authors: Bedru, Hayat , Zhao, Wenhong , Alrashoud, Mubarak , Tolba, Amr , Guo, He , Xia, Feng
- Date: 2020
- Type: Text , Journal article
- Relation: IEEE Access Vol. 8, no. (2020), p. 201015-201027
- Full Text:
- Reviewed:
- Description: Subnetwork identification plays a significant role in analyzing, managing, and comprehending the structure and functions in big networks. Numerous approaches have been proposed to solve the problem of subnetwork identification as well as community detection. Most of the methods focus on detecting communities by considering node attributes, edge information, or both. This study focuses on discovering subnetworks containing researchers with similar or related areas of interest or research topics. A topic- aware subnetwork identification is essential to discover potential researchers on particular research topics and provide qualitywork. Thus, we propose a topic-based optimal subnetwork identification approach (TOSNet). Based on some fundamental characteristics, this paper addresses the following problems: 1)How to discover topic-based subnetworks with a vigorous collaboration intensity? 2) How to rank the discovered subnetworks and single out one optimal subnetwork? We evaluate the performance of the proposed method against baseline methods by adopting the modularity measure, assess the accuracy based on the size of the identified subnetworks, and check the scalability for different sizes of benchmark networks. The experimental findings indicate that our approach shows excellent performance in identifying contextual subnetworks that maintain intensive collaboration amongst researchers for a particular research topic. © 2020 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.
- Authors: Bedru, Hayat , Zhao, Wenhong , Alrashoud, Mubarak , Tolba, Amr , Guo, He , Xia, Feng
- Date: 2020
- Type: Text , Journal article
- Relation: IEEE Access Vol. 8, no. (2020), p. 201015-201027
- Full Text:
- Reviewed:
- Description: Subnetwork identification plays a significant role in analyzing, managing, and comprehending the structure and functions in big networks. Numerous approaches have been proposed to solve the problem of subnetwork identification as well as community detection. Most of the methods focus on detecting communities by considering node attributes, edge information, or both. This study focuses on discovering subnetworks containing researchers with similar or related areas of interest or research topics. A topic- aware subnetwork identification is essential to discover potential researchers on particular research topics and provide qualitywork. Thus, we propose a topic-based optimal subnetwork identification approach (TOSNet). Based on some fundamental characteristics, this paper addresses the following problems: 1)How to discover topic-based subnetworks with a vigorous collaboration intensity? 2) How to rank the discovered subnetworks and single out one optimal subnetwork? We evaluate the performance of the proposed method against baseline methods by adopting the modularity measure, assess the accuracy based on the size of the identified subnetworks, and check the scalability for different sizes of benchmark networks. The experimental findings indicate that our approach shows excellent performance in identifying contextual subnetworks that maintain intensive collaboration amongst researchers for a particular research topic. © 2020 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.
Using links to aid web classification
- Xie, Wei, Mammadov, Musa, Yearwood, John
- Authors: Xie, Wei , Mammadov, Musa , Yearwood, John
- Date: 2007
- Type: Text , Conference paper
- Relation: Paper presented at 6th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2007, Melbourne, Victoria : 11th-13th July 2007 p. 981-986
- Full Text:
- Description: In this paper, we will present a new approach of using link information to improve the accuracy and efficiency of web classification. However, different from others, we only use the mappings between linked documents and their own class or classes. In this case, we only need to add a few features called linked-class features into the datasets. We apply SVM and BoosTexter for classification. We show that the classification accuracy can be improved based on mixtures of ordinary word features and out-linked-class features. We analyze and discuss the reason of this improvement.
- Description: 2003005438
- Authors: Xie, Wei , Mammadov, Musa , Yearwood, John
- Date: 2007
- Type: Text , Conference paper
- Relation: Paper presented at 6th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2007, Melbourne, Victoria : 11th-13th July 2007 p. 981-986
- Full Text:
- Description: In this paper, we will present a new approach of using link information to improve the accuracy and efficiency of web classification. However, different from others, we only use the mappings between linked documents and their own class or classes. In this case, we only need to add a few features called linked-class features into the datasets. We apply SVM and BoosTexter for classification. We show that the classification accuracy can be improved based on mixtures of ordinary word features and out-linked-class features. We analyze and discuss the reason of this improvement.
- Description: 2003005438
Venue topic model-enhanced joint graph modelling for citation recommendation in scholarly big data
- Wang, Wei, Gong, Zhiguo, Ren, Jing, Xia, Feng, Lv, Zhihan, Wei, Wei
- Authors: Wang, Wei , Gong, Zhiguo , Ren, Jing , Xia, Feng , Lv, Zhihan , Wei, Wei
- Date: 2021
- Type: Text , Journal article
- Relation: ACM Transactions on Asian and Low-Resource Language Information Processing Vol. 20, no. 1 (2021), p.
- Full Text:
- Reviewed:
- Description: Natural language processing technologies, such as topic models, have been proven to be effective for scholarly recommendation tasks with the ability to deal with content information. Recently, venue recommendation is becoming an increasingly important research task due to the unprecedented number of publication venues. However, traditional methods focus on either the author's local network or author-venue similarity, where the multiple relationships between scholars and venues are overlooked, especially the venue-venue interaction. To solve this problem, we propose an author topic model-enhanced joint graph modeling approach that consists of venue topic modeling, venue-specific topic influence modeling, and scholar preference modeling. We first model the venue topic with Latent Dirichlet Allocation. Then, we model the venue-specific topic influence in an asymmetric and low-dimensional way by considering the topic similarity between venues, the top-influence of venues, and the top-susceptibility of venues. The top-influence characterizes venues' capacity of exerting topic influence on other venues. The top-susceptibility captures venues' propensity of being topically influenced by other venues. Extensive experiments on two real-world datasets show that our proposed joint graph modeling approach outperforms the state-of-The-Art methods. © 2020 ACM.
- Authors: Wang, Wei , Gong, Zhiguo , Ren, Jing , Xia, Feng , Lv, Zhihan , Wei, Wei
- Date: 2021
- Type: Text , Journal article
- Relation: ACM Transactions on Asian and Low-Resource Language Information Processing Vol. 20, no. 1 (2021), p.
- Full Text:
- Reviewed:
- Description: Natural language processing technologies, such as topic models, have been proven to be effective for scholarly recommendation tasks with the ability to deal with content information. Recently, venue recommendation is becoming an increasingly important research task due to the unprecedented number of publication venues. However, traditional methods focus on either the author's local network or author-venue similarity, where the multiple relationships between scholars and venues are overlooked, especially the venue-venue interaction. To solve this problem, we propose an author topic model-enhanced joint graph modeling approach that consists of venue topic modeling, venue-specific topic influence modeling, and scholar preference modeling. We first model the venue topic with Latent Dirichlet Allocation. Then, we model the venue-specific topic influence in an asymmetric and low-dimensional way by considering the topic similarity between venues, the top-influence of venues, and the top-susceptibility of venues. The top-influence characterizes venues' capacity of exerting topic influence on other venues. The top-susceptibility captures venues' propensity of being topically influenced by other venues. Extensive experiments on two real-world datasets show that our proposed joint graph modeling approach outperforms the state-of-The-Art methods. © 2020 ACM.
Visual tools for analysing evolution, emergence, and error in data streams
- Hart, Sol, Yearwood, John, Bagirov, Adil
- Authors: Hart, Sol , Yearwood, John , Bagirov, Adil
- Date: 2007
- Type: Text , Conference paper
- Relation: Paper presented at 6th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2007, Melbourne, Victoria : 11th-13th July 2007 p. 987-992
- Full Text:
- Description: The relatively new field of stream mining has necessitated the development of robust drift-aware algorithms that provide accurate, real time, data handling capabilities. Tools are needed to assess and diagnose important trends and investigate drift evolution parameters. In this paper, we present two new and novel visualisation techniques, Pixie and Luna graphs, which incorporate salient group statistics coupled with intuitive visual representations of multidimensional groupings over time. Through the novel representations presented here, spatial interactions between temporal divisions can be diagnosed and overall distribution patterns identified. It provides a means of evaluating in non-constrained capacity, commonly constrained evolutionary problems.
- Description: 2003005432
- Authors: Hart, Sol , Yearwood, John , Bagirov, Adil
- Date: 2007
- Type: Text , Conference paper
- Relation: Paper presented at 6th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2007, Melbourne, Victoria : 11th-13th July 2007 p. 987-992
- Full Text:
- Description: The relatively new field of stream mining has necessitated the development of robust drift-aware algorithms that provide accurate, real time, data handling capabilities. Tools are needed to assess and diagnose important trends and investigate drift evolution parameters. In this paper, we present two new and novel visualisation techniques, Pixie and Luna graphs, which incorporate salient group statistics coupled with intuitive visual representations of multidimensional groupings over time. Through the novel representations presented here, spatial interactions between temporal divisions can be diagnosed and overall distribution patterns identified. It provides a means of evaluating in non-constrained capacity, commonly constrained evolutionary problems.
- Description: 2003005432