Fake Profile Detection Using Machine Learning Techniques

Abstract

Our lives are significantly impacted by social media platforms such as Facebook, Twitter, Instagram, LinkedIn, and others. People are actively participating in it the world over. However, it also has to deal with the issue of bogus profiles. False accounts are frequently created by humans, bots, or computers. They are used to disseminate rumors and engage in illicit activities like identity theft and phishing. So, in this project, the author'll talk about a detection model that uses a variety of machine learning techniques to distinguish between fake and real Twitter profiles based on attributes like follower and friend counts, status updates, and more. The author used the dataset of Twitter profiles, separating real accounts into TFP and E13 and false accounts into INT, TWT, and FSF. Here, the author discusses LSTM, XG Boost, Random Forest, and Neural Networks. The key characteristics are chosen to assess a social media profile's authenticity. Hyperparameters and the architecture are also covered. Finally, results are produced after training the models. The output is therefore 0 for genuine profiles and 1 for false profiles. When a phony profile is discovered, it can be disabled or destroyed so that cyber security problems can be prevented. Python and the necessary libraries, such as Sklearn, Numpy, and Pandas, are used for implementation. At the end of this study, the author will come to the conclusion that XG Boost is the best machine learning technique for finding fake profiles.

Fake account detection in social media using machine learning methods: literature review

“…Other platforms are excluded from this literature review. 1,162 accounts Gupta and Kaushal [7] 4,708 accounts Khalil et al [19] Fake accounts: 13,000 Real accounts: 5,386 Twitter Ersahin et al [8] Fake accounts: 501 Real accounts: 499 Cresci et al [18] 13,101 accounts Walt and Eloff [20] 223,796 accounts Akyon and Kalfaoglu [24] Fake accounts: 700 Real accounts: 700 Bharti and Pandey [33] Real accounts: 1,103 Narayan [34] Fake accounts: 1,056 Real accounts: 1,176 Instagram Meshram et al [14] Fake accounts: 3,231 Real accounts: 6,868 Purba et al [15] Fake accounts: 32,869 Real accounts: 32,460 Sheikhi [1] Fake accounts: 3,132 Real accounts: 6,868 Durga and Sudhakar [40] Fake accounts: 201 Real accounts: 1,002 [29] The fake project dataset 11,737 accounts Khaled et al [26] MIB dataset Fake accounts: 3,351 Real accounts: 1,950 Wang et al [31] CLEF2019 dataset 7,120 accounts Bharti and Pandey [33] The fake project [18] 5.870 accounts Chakraborty et al [36] MIB dataset Fake accounts: 3,474 Real accounts: 3,351 Kadam and Sharma [38] GitHub 2,820 accounts Instagram Kesharwani et al [32] Fake, spammer, and genuine Instagram accounts 696 accounts Das et al [37] Kaggle dataset 576 accounts…”

Profile Verification and Secured Social Engineering with Machine Learning Models

“…These can belong to various domains such as supervised machine learning, artificial neural networks, regression models, etc. Partha Chakraborty et al present in their report a few of these algorithms, namely Random Forest Classifier, Long Short Term Memory [5,9] Algorithm, and Extreme Gradient Booth, with a resultant discussion on Neural Networks. The following also introduces the Sybil Rank method that uses a single node along with its limitations [9].…”

“…The first step for the development and deployment of an efficient machine learning-based model is to understand the viability of input data, along with the need for preprocessing. Based on the profiling of the current online social networks (OSNs), there have been found certain markers in the form of following volume [3,4,9], description length, post activity, listing counts statuses, etc. As cited in the 2022 study [9], the graphical representation of this tabulated data helps establish the classification and categorization of such accounts as legitimate or fake.…”

“…Based on the profiling of the current online social networks (OSNs), there have been found certain markers in the form of following volume [3,4,9], description length, post activity, listing counts statuses, etc. As cited in the 2022 study [9], the graphical representation of this tabulated data helps establish the classification and categorization of such accounts as legitimate or fake. This dataset often presents the underlying data type as string, float, and integer; with categorization as Boolean variables of "true" and "False".…”

“…The use of CSV file type as appropriate input training and testing dataset can be weighted based on the type conversions and available functionalities as mentioned in [9,11]. To transform these into processable formats, one can utilize the DataFrame concept of Arrays in NumPy.…”