Close Menu
DISADISA
  • Home
  • News
  • Social Media
  • Disinformation
  • Fake Information
  • Social Media Impact
Trending Now

Citizen Journalism in the Emmanuel Case: A Double-Edged Sword of News Dissemination and Misinformation

September 4, 2025

Political Scapegoating of Artificial Intelligence in the Disinformation Era

September 4, 2025

US Health Secretary RFK Jr. Faces Calls for Resignation Over Misinformation

September 4, 2025
Facebook X (Twitter) Instagram
Facebook X (Twitter) Instagram YouTube
DISADISA
Newsletter
  • Home
  • News
  • Social Media
  • Disinformation
  • Fake Information
  • Social Media Impact
DISADISA
Home»Fake Information»Ensemble Learning for Detecting Fake Instagram Profiles
Fake Information

Ensemble Learning for Detecting Fake Instagram Profiles

Press RoomBy Press RoomSeptember 4, 2025
Facebook Twitter Pinterest LinkedIn Tumblr Email

Unmasking Deception: A Novel Hybrid Model for Enhanced Fake Profile Detection on Instagram

The proliferation of fake profiles on social media platforms, particularly Instagram, poses a significant threat to online security and user trust. These profiles are often used to spread misinformation, spam, and malicious content, eroding the integrity of online communities. To combat this growing problem, researchers have developed a novel hybrid machine learning model that demonstrates exceptional accuracy in identifying fake Instagram accounts. This innovative approach combines the strengths of established algorithms like Random Forest (RF) and XGBoost, optimized with cutting-edge techniques to achieve unprecedented detection capabilities.

The model’s foundation lies in a carefully orchestrated workflow that begins with data retrieval and merging, followed by rigorous pre-processing steps crucial for handling the intricacies of Instagram datasets. Hyperparameter tuning using GridSearchCV optimizes the Random Forest algorithm, meticulously balancing depth and tree count for optimal performance and generalization. Simultaneously, XGBoost is optimized using scale_pos_weight to address class imbalance, a common challenge in fake profile detection, without the computational overhead of oversampling. This strategic combination of optimized RF and XGBoost, through weighted voting, leverages the unique strengths of each algorithm: RF’s proficiency in handling structured data and XGBoost’s gradient boosting capabilities.

The model further incorporates meticulous feature engineering to extract relevant profile attributes and behavioral patterns indicative of fake accounts. Features such as username length, biography length, follower count, following count, and activity metrics are carefully analyzed and integrated into the model. Rigorous evaluation using cross-validation ensures the model’s consistency and reliability. The result is a hybrid model that achieves a remarkable 98.24% accuracy, significantly outperforming individual classifiers.

The innovation lies not just in employing existing algorithms but in their strategic optimization and combination within a hybrid framework. This framework incorporates enhanced class balancing techniques, tailored feature extraction strategies, and improved decision-making processes, leading to a more reliable and interpretable fake profile detection model. The systematic testing and analysis provide valuable insights into the model’s behavior, resilience, and adaptability.

The mathematical underpinnings of the model are detailed in the research, providing a step-by-step illustration of the data pre-processing, model optimization, and fusion processes. Equations outline the SMOTE implementation for class imbalance handling, hyperparameter tuning for RF and XGBoost, and the weighted voting mechanism for the final prediction. Algorithm “InstaFake” further elucidates the model’s procedure, encompassing data balancing, training-testing split, classifier selection, and prediction mechanisms. This rigorous mathematical framework provides a transparent and reproducible approach to fake profile detection.

The performance evaluation utilizes a confusion matrix to provide a clear visualization of the model’s accuracy in distinguishing between real and fake profiles. The model achieved remarkable precision, with 207 correctly identified fake profiles (true positives) and 184 correctly identified real profiles (true negatives). The minimal false positives and zero false negatives demonstrate the model’s high reliability. Key performance metrics, including recall, precision, accuracy, and F1-score, underscore the model’s efficacy. The model achieves a 100% recall for fake profiles, indicating its ability to capture all fraudulent accounts, and a 96% specificity for real profiles, signifying its precision in identifying genuine accounts. The overall accuracy of 98% and F1-score of 98% highlight the balanced performance and robustness of the model.

Comparative analysis against baseline models such as Logistic Regression, SVM, and Decision Tree, all utilizing SMOTE for class balancing, demonstrates the superior performance of the proposed hybrid model. Further comparisons with existing models in the literature reinforces the model’s state-of-the-art performance, showcasing its potential to significantly improve online safety and user trust. The research also acknowledges limitations, such as the relatively small dataset size, and suggests directions for future research to enhance the model’s generalizability and address evolving fake profile tactics. This work represents a substantial contribution to the field of social media security, offering a promising solution to the pervasive problem of fake profiles on Instagram.

Share. Facebook Twitter Pinterest LinkedIn WhatsApp Reddit Tumblr Email

Read More

Fake News Detection and Classification with Transfer Learning and Large Language Models

September 4, 2025

Social Media Bans Rejected: User Information Filtering Capabilities Mitigate Concerns (2019 Data from Statista)

September 3, 2025

Fraudulent Social Media Accounts Exploit Kiingitanga Followers.

September 2, 2025

Our Picks

Political Scapegoating of Artificial Intelligence in the Disinformation Era

September 4, 2025

US Health Secretary RFK Jr. Faces Calls for Resignation Over Misinformation

September 4, 2025

Presidential Condemnation of AI: A New Form of Misinformation?

September 4, 2025

Deputy Chief Minister of Tamil Nadu Highlights Disinformation and Misinformation as Emerging Global Threats

September 4, 2025
Stay In Touch
  • Facebook
  • Twitter
  • Pinterest
  • Instagram
  • YouTube
  • Vimeo

Don't Miss

News

College Students and Misinformation: An Archival Analysis from The Sun Times News

By Press RoomSeptember 4, 20250

The Misinformation Epidemic: How College Students Navigate a Sea of Falsehoods College campuses, once seen…

Ensemble Learning for Detecting Fake Instagram Profiles

September 4, 2025

Contesting Disinformation: A Discussion with David Frum

September 4, 2025

Apeel Sciences Initiates Legal Action Against Influencer to Combat Misinformation Regarding Produce Preservation Technology

September 4, 2025
DISA
Facebook X (Twitter) Instagram Pinterest
  • Home
  • Privacy Policy
  • Terms of use
  • Contact
© 2025 DISA. All Rights Reserved.

Type above and press Enter to search. Press Esc to cancel.