Instagram Spam Detection (ISD)

Pranali V., Dhote and Rima, Ramteke and Anjali R., Patel and Pratik, Dewalker and Prof. Anupam, Chaube (2024) Instagram Spam Detection (ISD). International Journal of Trend in Scientific Research and Development, 8 (5). pp. 573-583. ISSN 2456-6470

Text ijtsrd69419.pdf Download (1MB)

Abstract

An Instagram spam detection project would cover the project's aim, methods, and outcomes. It would detail how the project identifies and filters out spam content on Instagram to enhance user experience and security. The abstract might mention the use of machine learning algorithms, natural language processing, and image recognition to detect and remove spam posts, comments, and accounts. It could also highlight the importance of such a project in maintaining a clean and authentic environment on the platform. The focus is on developing algorithms and tools to automatically identify and filter out spam content on the platform. This involves using machine learning techniques to analyze patterns in user behavior, text content, and images to distinguish between legitimate posts and spam. By training the system on a large dataset of known spam content, the model can learn to recognize and flag suspicious activity in real-time, helping to maintain a safe and enjoyable environment for user. Assuring a safe and fulfilling user experience on the well-known social media platform requires the crucial duty of Instagram Spam Detection (ISD). Cyberbullying, identity theft, and even financial fraud can result from unsolicited messages on Instagram. Several machine learning algorithms for Instagram spam detection have been presented by researchers as a solution to this problem. Tokenization, stop word removal, and sentiment analysis using the VADER algorithm are some methods of preparing the text data. Following preprocessing, Count Vectorizer is used to turn the data into numerical feature vectors. Based on the labeled data, three classifiers are trained and assessed: F1 score, accuracy, precision, recall, and Decision Trees/Random Forest. Taking into consideration weighted parameters that are essential in establishing an account's legitimacy, Gradient Boosting Classifier has demonstrated encouraging results in detecting phony accounts on Instagram. Instagram streaming spam detection continues to be difficult, and a strong detection method should take into account the elements of popular topics, content, URL, and user identity.

Item Type:	Article
Subjects:	T Technology > T Technology (General)
Divisions:	Postgraduate > Master's of Islamic Education
Depositing User:	Journal Editor
Date Deposited:	23 Oct 2024 09:36
Last Modified:	23 Oct 2024 09:36
URI:	http://eprints.umsida.ac.id/id/eprint/14233

Actions (login required)

View Item