Brunel Elvire BOUYA-MOKO

Mr. Brunel Elvire BOUYA-MOKO

Jiangsu University, Zhenjiang, Jiangsu, China
+1
orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 May 2021 in Computers & Security
Verified
Journal Article
Closed access
Edward Kwadwo BoahenEdward Kwadwo Boahen
Edward Kwadwo Boahen
Brunel Elvire BOUYA-MOKOBrunel Elvire BOUYA-MOKO
Brunel Elvire BOUYA-MOKO
CWChangda Wang
Changda Wang

2 September 2022 in IEEE Transactions on Computational Social Systems
Verified
Journal Article
Closed access
Edward Kwadwo BoahenEdward Kwadwo Boahen
Edward Kwadwo Boahen
Brunel Elvire BOUYA-MOKOBrunel Elvire BOUYA-MOKO
Brunel Elvire BOUYA-MOKO
Faizan QamarFaizan Qamar
Faizan Qamar
CWChangda Wang
Changda Wang
The major threat to online social network (OSN) users is account compromisation. Spammers now spread malicious messages by exploiting the trust relationship established between account owners and their friends. The challenge in detecting a compromised account by service providers is validating the trusted relationship established between the account owners, their friends, and the spammers. Another challenge is the increase in required human interaction with feature selection. Research available on supervised learning has limitations with feature selection and accounts that cannot be profiled, like application programming interface (API). Therefore, this article discusses the various behaviors of OSN users and the current approaches in detecting a compromised OSN account, emphasizing its limitations and challenges. We propose a deep learning approach that addresses and resolve the constraints faced by previous schemes. We detailed our proposed optimized nonsymmetric deep autoencoder (OPT $\_$ NSDAE) for unsupervised feature learning, which reduces the required human interaction levels in the selection and extraction of features. We evaluated our proposed classifier using three different social network datasets, Facebook, Google $+$ , and Twitter, in addition to the NSL-KDD and KDDCUP’99 datasets, in a graphical-user-interface-enabled Weka application. The experimental results show that our proposed approach outperformed most of the traditional schemes in OSN compromised account detection.

29 June 2020 in Applied Artificial Intelligence
Verified
Journal Article
Closed access
Edward Kwadwo BoahenEdward Kwadwo Boahen
Edward Kwadwo Boahen
CWChangda Wang
Changda Wang
Brunel Elvire BOUYA-MOKOBrunel Elvire BOUYA-MOKO
Brunel Elvire BOUYA-MOKO
The primary threat to online social network (OSN) users is account compromisation. The challenge in detecting a compromised account is due to the trusted relationship established between the account owners, their friends, and the service providers. The available research which focuses on using machine learning has limitations with human experts involved in feature selection and a standardized dataset. The paper discusses users` various behaviors of users of OSN and the up-to-date approaches in detecting a compromised OSN account with emphasis on the limitations and challenges. Furthermore, we propose an enhanced machine learning approach Word Embedding and KNN (WE-KNN), which addresses the limitations faced by the previous techniques used. We detailed our proposed WE-KNN for feature extraction, selection of behavior of OSN users, and classification. Our proposed model is evaluated using the standard benchmark datasets, namely KDD Cup ‘99 and NSL-KDD and implemented it in WEKA. Besides, we used state-of-the-art evaluation metrics to assess the performance of our model. The results obtained depicts that the proposed approach in compromise account detection performs better.

25 May 2022 in Electronics
Verified
Journal Article
Open access
PDF file
Brunel Elvire BOUYA-MOKOBrunel Elvire BOUYA-MOKO
Brunel Elvire BOUYA-MOKO
Edward Kwadwo BoahenEdward Kwadwo Boahen
Edward Kwadwo Boahen
CWChangda Wang
Changda Wang
Strong network connections make the risk of malicious activities emerge faster while dealing with big data. An intrusion detection system (IDS) can be utilized for alerting suitable entities when hazardous actions are occurring. Most of the techniques used to classify intrusions lack the techniques executed with big data. This paper devised an optimization-driven deep learning technique for detecting the intrusion using the Spark model. The input data is fed to the data partitioning phase wherein the partitioning of data is done using the proposed fuzzy local information and Bhattacharya-based C-means (FLIBCM). The proposed FLIBCM was devised by combining Bhattacharya distance and fuzzy local information C-Means (FLICM). The feature selection was achieved with classwise info gained to select imperative features. The data augmentation was done with oversampling to make it apposite for further processing. The detection of intrusion was done using a deep Maxout network (DMN), which was trained using the proposed student psychology water cycle caviar (SPWCC) obtained by combining the water cycle algorithm (WCA), the conditional autoregressive value at risk by regression quantiles (CAViaR), and the student psychology-based optimization algorithm (SPBO). The proposed SPWCC-based DMN offered enhanced performance with the highest accuracy of 97.6%, sensitivity of 98%, and specificity of 97%.