Probabilistic graphical model for detecting spammers in microblog websites

Han, Zhongming; Xu, Fengmin; Duan, Dagao

doi:10.1504/ijes.2016.073747

Cited by 4 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The SVM classifier establishes a learning mode that assigns new instances to a non-probabilistic binary linear classifier. Taking NSL-KDD Cup 99 as the experimental dataset, a wrapper feature selection algorithm is proposed based on SVM (Han et al, 2017). The experimental results show that the use of fewer feature attributes can achieve a high classification accuracy of 91% in the training set and 99% of the classification accuracy with 36 attributes.…”

Section: Literature Surveymentioning

confidence: 99%

A combination classification method based on Ripper and Adaboost

Wang¹,

Chen²,

Zhang³

et al. 2021

IJES

View full text Add to dashboard Cite

With the growing demand for data analysis, machine learning technology has been widely used in many applications, such as mass data summarising rules, predicting behaviours and dividing characteristics. The Ripper algorithm presents better pruning and stopping criteria than the traditional decision tree algorithm (C4.5), while its error rate less than or equal to C4.5 by O(nlog2n) time complexity. As a result of that, Ripper can maintain high efficiency even on the massive dataset which contains lots of noise. Adaboost is one of iterative algorithms, which combines a group of weak classifiers together to set up a strong classifier. In order to improve the accuracy of Ripper classification algorithm and reduce the computational complexity, this paper proposes a Ripper-Adaboost combined classification method (Ripper-ADB). The experiment result shows Ripper-ADB could improve the classifier and get higher classification accuracy than decision tree and SVM.

show abstract

Section: Literature Surveymentioning

confidence: 99%

A combination classification method based on Ripper and Adaboost

Wang¹,

Chen²,

Zhang³

et al. 2021

IJES

View full text Add to dashboard Cite

show abstract

“…Another work (Cardoso et al, 1999) integrates uncertainty reasoning using possibility logic with Petri nets, which was more elaborated in Lee et al (2003). Furthermore, and even aside from diagnostic purposes, probabilistic reasoning is being used in a variety of fields, such as cyber-security (Xu et al, 2016;Han et al, 2016), performance modelling (Zhang et al, 2016) and prediction (Lu et al, 2016).…”

Section: Related Workmentioning

confidence: 99%

Distributed diagnosis based on distributed probability propagation nets

Berghout

Bennoui

2019

IJCSE

View full text Add to dashboard Cite

This paper addresses the problem of modelling uncertainty in the distributed context. It is situated in the field of diagnosis; more precisely, model-based diagnosis of distributed systems. A special focus is given to modelling uncertainty and probabilistic reasoning. Thus, this work is based on a probabilistic modelling formalism called: 'probability propagation nets' (PPNs), which are designed for centralised systems. Hence, an extension of this model is proposed to suit the distributed context. Distributed probability propagation nets (DPPNs), the proposed extension, were conceived to consider the distributed systems' particularities. So, the set we consider is a set of interacting subsystems, each of which is modelled by a DPPN. The interaction among the subsystems is modelled through the firing of common transitions belonging to more than one subsystem. All of that is logically supported by means of probabilistic Horn abductions (PHAs). Furthermore, the diagnostic process is done by exploiting transition-invariants, a diagnostic technique developed for Petri nets. The proposed extension is illustrated through a real life example.

show abstract

“…In addition, spam detection work in Web 2.0 also has close and significant inspiration for IWA identification. Behavior-based work can be further divided into two categories: direct calculation type [1][2] and training-leaning type [3]. In the direct calculation type, Chen Kai et al [1] aimed at the account list which has forwarded hot topic.…”

Section: Related Workmentioning

confidence: 99%

“…However, in this method, the weight of dimensions is set relatively subjective. In the leaning-training type, Han Zhongming et al [3] transform the behavior and features of the micro-blog users to characteristic vector, constructing a probabilistic graph of features and behavior:…”

Section: Related Workmentioning

confidence: 99%

A DBN-Based Classifying Approach to Discover the Internet Water Army

Sun

Zhao

Niu

et al. 2014

Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications

View full text Add to dashboard Cite

Abstract. The Internet water army (IWA) usually refers to hidden paid posters and collusive spammers, which has already generated big threats for cyber security. Many researchers begin to study how to effectively identify the IWA. Currently, most efforts to distinguish non-IWA and IWA in data mining context focus on utilizing classification-based algorithms, including Bayesian Network, SVM, KNN and etc... However, Bayesian Network need strong conditional independence assumption, KNN has big computation costs, above approach may affect the effectiveness to some extent in real industrial applications. Hence, Neural Networks-like deep approach for IWA identification gradually becomes an emerging but possible direction and attempt. Unfortunately, there also exists one main problem, which is how to balance the deep learning and computation costs in hierarchical architecture. More specially, combine leaning-level heuristic training design and computing-level concurrent computation is a challenging issue. In this paper, we propose a collaborative hierarchical approach based on the deep belief network (DBN) for IWA identification. Firstly, a DBN-based collaborative model with hierarchical classifying mechanism is built. Then towards Hadoop platform, the Downpour Stochastic gradient descent (Downpour SGD) is exploited for DBN pre-training. Finally, the dynamical workflow will be designed for managing the whole learning-based classifying process. The experimental evaluation shows that the valid of our approach.

show abstract

Probabilistic graphical model for detecting spammers in microblog websites

Cited by 4 publications

References 0 publications

A combination classification method based on Ripper and Adaboost

A combination classification method based on Ripper and Adaboost

Distributed diagnosis based on distributed probability propagation nets

A DBN-Based Classifying Approach to Discover the Internet Water Army

Contact Info

Product

Resources

About