Identifying Spam Accounts on Instagram: An Analysis of User Activity Data Using Machine Learning

Instances of the spam bots involving social networks such as Instagram pose a huge threat to the functionality and authenticity of such platforms. This study analyses the detection accuracy of accepted methods when dealing with spam accounts. Through this analysis, authors focused on compare the pros and cons of each method and select the approach that suits their needs. The purpose of this study is the assessment of the efficacy of an ensemble learning method for the detection of spam accounts on Instagram. At the next stage, a novel approach is discussed that unite some features of the highly functioning methods. With the help of user activity data, which will be used to train and test the machine learning approach, we strive to achieve significantly better results in spam account detection compared to existing methods. Our approach focusses on conducting the research details i.e. how the data collection and tabulation for analysis will take place. However, this involves making data clean, filling in missing information, and possibly even balancing lopsided datasets. Discovery of features is a very crucial step which will help us identify worthful details from user activity data including how often user posts, how users content perform and account characteristics. The evaluation metrics focus on precision, recall, and F1-score to assess the model's ability and accurately classify both spam and genuine accounts. The results demonstrate that the ensemble model achieves a high degree of accuracy, with precision exceeding 0.96 and recall exceeding 0.95.