A PSO-Based Feature Subset Selection for Application of Spam /Non-spam Detection

Behjat, Amir; Nezamabadi‐pour, Hossein; Sulaiman, Md. Nasir; Mustapha, Norwati

doi:10.1007/978-3-642-40567-9_16

Cited by 8 publications

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Results indicated that the employment of such mechanisms can improve the accuracy performance. In addition, Zhang et al [25] and Behjat et al [26] introduced the Particle Swarm Optimization (PSO) into the Internet water army identification problem. These evaluable achievements to a large extent increase the detection accuracy of Internet water armies.…”

Section: Related Workmentioning

confidence: 99%

An Internet Water Army Detection Supernetwork Model

et al. 2019

View full text Add to dashboard Cite

The emergence of Internet water armies has strongly affected the information quality of online communication platforms, thus disrupting the order of the Internet. Accurate detection of Internet water armies is therefore of great significance. Based on the supernetwork theory, a new Internet water army detection model is proposed in this paper, in which a supernetwork with four layers is established, including social subnetwork, information subnetwork, psychological subnetwork, and negative keyword subnetwork. Then, personal information of users, dissemination process of information, the transformation process of different psychologies, the similarity between different keywords, and the connections between different subnetworks are considered in the model. Thus, nine composite indexes are proposed, the majority of which are used for the first time in detecting Internet water armies. A dataset selected from the largest online communication platform in China, the Weibo website, is used to test the performance of the model. Four existing water army detection models introduced in previous studies are used to provide a comparison analysis. The results show that our proposed model has better performance in terms of accuracy and stability than the other four existing models, which thanks to the employment of the supernetwork theory. We believe that our proposed model could be helpful for information researchers to further understand the complex nature of Internet water armies, as well as for the government to better manage the Internet.

show abstract

Section: Related Workmentioning

confidence: 99%

An Internet Water Army Detection Supernetwork Model

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Behjat et al [43] used Binary Particle Swarm Optimization as feature selection and Multi-Layer Perceptron as classification and bag of word as extraction; they compare with information gain as feature selection and three classifiers, namely, BP Neural Network, Linear Discriminant and support vector machine. Their proposed output from genetic algorithm and other feature selection of information gain are the feature selection and other classifiers like BP Neural Network, Linear Discriminant and support vector machine.…”

Section: B Supervised Feature Selectionsmentioning

confidence: 99%

“…• The researchers assumed that feature selection can decrease the number of features [23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40] in carrying out a comparison between "Harmony Search, genetic algorithm and others mentioned before" of feature selection without detecting the weaknesses of these algorithms for Islamic terrorists' web. They offered a solution for such weaknesses by hybrid [41][42][43][44] with other algorithms and then compared the applied algorithms with another one so as to increase their performance, without addressing the effect with other processes in the web classifiers such as classifiers algorithms and extraction method.…”

Section: Limitations Of Feature Extraction and Term Extractionmentioning

confidence: 99%

A Systematic Review of Islamic Terrorist Webpages and Keywords

Al-Sukhni¹,

Saudi²,

Ahmad³

2019

MJoSHT

View full text Add to dashboard Cite

As the internet is considerably expanding and on-line information is increasing, the identification and detection of the large amounts of different web information become vitally important, particularly those of dark web or Islamic extremists. Webpages with extremist and terrorist content are believed to be the main factors in the radicalisation and recruitment of disaffected individuals who might be involved in terrorist activities at home or those who fight alongside terrorist groups abroad. In fact, the sheer volume of online data makes it practically impossible for authorities to carry out the individual examination for every webpage, post or conversational thread that might or might not be relevant to terrorism or contain terrorist sympathies. As terrorists exist within every nation and every religion, hence this paper presents a review and systematic analysis of existing webpages on Islamic terrorists. This include existing database of Islamic extremist words and existing techniques of web classifier for keywords. Based on this systematic review analysis, it will be the input to the formation of a new Islamic extremists WorldNet.

show abstract

“…MLPNN) was proposed for spam email identification in[16].PSO algorithm is used for characteristic selection and MLPNN model is used for training, data testing. In PSO-MLPNN model the perceptron neural function is used with sigmoid activation function for the hidden layer; %80 of the data is used for training and %20 for testing.…”

mentioning

confidence: 99%

Detection of Spam Email by Combining Harmony Search Algorithm and Decision Tree

Gashti

2017

Eng. Technol. Appl. Sci. Res.

View full text Add to dashboard Cite

Spam emails is probable the main problem faced by most e-mail users. There are many features in spam email detection and some of these features have little effect on detection and cause skew detection and classification of spam email. Thus, Feature Selection (FS) is one of the key topics in spam email detection systems. With choosing the important and effective features in classification, its performance can be optimized. Selector features has the task of finding a subset of features to improve the accuracy of its predictions. In this paper, a hybrid of Harmony Search Algorithm (HSA) and decision tree is used for selecting the best features and classification. The obtained results on Spam-base dataset show that the rate of recognition accuracy in the proposed model is 95.25% which is high in comparison with models such as SVM, NB, J48 and MLP. Also, the accuracy of the proposed model on the datasets of Ling-spam and PU1 is high in comparison with models such as NB, SVM and LR.

show abstract

A PSO-Based Feature Subset Selection for Application of Spam /Non-spam Detection

Cited by 8 publications

References 18 publications

An Internet Water Army Detection Supernetwork Model

An Internet Water Army Detection Supernetwork Model

A Systematic Review of Islamic Terrorist Webpages and Keywords

Detection of Spam Email by Combining Harmony Search Algorithm and Decision Tree

Contact Info

Product

Resources

About