F2PMSMD: Design of a Fusion Model to Identify Fake Profiles from Multimodal Social Media Datasets

Aditya, B.; Gnanajeyaraman, R.; Hole, Shreyas Rajendra; Mohanty, Sachi Nandan

doi:10.1007/978-3-031-35081-8_2

Cited by 5 publications

(2 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, the popularity of the DL method in computer vision has become a spike of research interest and exploring DL for addressing the issues of spoof image detection activities [17]. Aditya et al (2023) introduced a novel fusion model, F2PMSMD, designed to identify fake profiles in multimodal social media datasets. The model collects various user information, such as profile pictures, username characteristics, and activity metrics, and uses a Genetic Algorithm (GA) for feature selection.…”

Section: Introductionmentioning

confidence: 99%

“…It employs a combination of classifiers including Naïve Bayes, Multilayer Perceptron, Logistic Regression, Support Vector Machine, and Deep Forest. The F2PMSMD model achieves impressive performance with 98.5% accuracy, 94.3% precision, 94.9% recall, and a 94.7% Fmeasure, making it suitable for deployment on various social media platforms to detect fake profiles [24]. Mangla et al (2023) present a low-cost EyeWriter system designed as an assisting aid for individuals who are unable to communicate verbally.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Geo-Tagged Spoofing Detection using Jaccard Similarity

Koparde,

Mane

2023

EAI Endorsed Trans Energy Web

View full text Add to dashboard Cite

In recent years, position evaluation of mobile devices has developed as an essential part of social movement. Meantime, the criminals may interfere with the information of geographical position (geo-position), and they can adjust the geo-position for their convenience. Therefore, it is important to identify the authenticity of geo-position. In this paper, an instant messaging platform-based geo-tagged spoof image detection system is created using Jaccard similarity. With the help of a Fuzzy filter, the input, as well as spoofing images, are subjected to camera footprint extraction, and their corresponding outputs are fused by Dice Coefficient. Moreover, the input as well as spoofed images is subjected to geotagged process, and their corresponding geotagged input, and geotagged spoofed images are fused by Tanimoto similarity. At last, the fused images from Dice Coefficient, and Tanimoto similarity are employed for the spoof detection process, where the Jaccard similarity compares the two images using Dicerete Cosine Transform (DCT). Consequently, the spoofed images are detected, and their effectiveness is measured in terms of accuracy, False Positive Rate (FPR), and True Positive Rate (TPR), as well as the corresponding values are attained like 0.099, 0.892, and 0.896 respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%