AI summary · arXiv: 2104.13094v1 · cs.LG

Detection of Fake Users in SMPs Using NLP and Graph Embeddings

Manojit Chakraborty, Shubham Das, Radhika Mamidi
Original paper · DOI 10.1109/comsnets53615.2022.9668371 Submitted to arXiv on 27 Apr. 2021 Published in 2022 14th International Conference on COMmunication Systems & NETworkS (COMSNETS) (04 Jan. 2022) 5 pages, 3 figures License: CC BY 4.0

How to read this summary? Illustrated guide

The essentials

  • Research focuses on detecting fake and spam user accounts on Twitter
  • Combination of Graph Representation Learning and Natural Language Processing techniques used
  • Large user base of social media platforms generates massive amount of data
  • Fake and spam accounts used by organizations for competitive advantage
  • Model's performance compared with other existing works in Table 2
  • Dataset size affects model's performance, older models rely on Twitter user IDs from 2018 or earlier
  • Researchers' database consists of 46k active Twitter accounts that passed Twitter's spam detection tool
  • Model achieves approximately 97% accuracy score, significant improvement over baseline
  • Approach utilizes optimized user metadata attributes, NLP features, and graph representation of user accounts
  • Proposed feature set demonstrates consistent performance across different models and class labels
  • Spam users tend to form communities by selectively following other users
  • Future studies will explore this observation further and investigate differences between legitimate human users vs. legitimate social bots, as well as human spammers vs. social bot spammers.
  • Impact of increased tweet length on automated spam accounts' ability to generate intelligent content needs further study due to recent changes in maximum tweet length limits.
  • Approach can be applied not only to Twitter but also to other social media platforms or real-time filtering applications for data collection pipelines.
  • Study provides valuable insights and lays foundation for future research in this field.

Key points written by AI: they may contain errors, check them against the paper.

AI-generated keywords: Fake and Spam User Detection

Read further

  • Comprehensive Summary
  • Layman's Summary
  • Blog article
  • Original abstract
This research focuses on detecting fake and spam user accounts on Twitter using a combination of Graph Representation Learning and Natural Language Processing techniques. The study acknowledges the large user base of social media platforms like Facebook, Twitter, and Instagram, which generates a massive amount of data every second. Many organizations use fake and spam accounts to gain a competitive edge over others. The authors compare their model's performance with other existing works in Table 2. They note that as the dataset size increases, the model's performance tends to decrease. Most models in Table 2 rely on Twitter user IDs from 2018 or earlier when there was a surge in fake, spam, and bot accounts. However, since mid-2018, Twitter has implemented robust spam detection and deletion technologies to eliminate most of these accounts. The researchers emphasize that their database consists of 46k active Twitter accounts that have successfully passed Twitter's spam detection tool. These accounts are much harder to detect as spam compared to older databases used by other models. Despite this challenge, their model achieves an accuracy score of approximately 97%, which is a significant improvement over the baseline. In conclusion, this work presents a novel and effective method for detecting spam users on social media platforms. The approach utilizes optimized user metadata attributes, features generated from NLP techniques, and graph representation of user accounts. The proposed feature set demonstrates consistent performance across different models and class labels. The researchers also observe that spam users tend to form communities by selectively following other users. They plan to explore this observation further in future studies. Additionally, they highlight the need for investigating the clear differences between legitimate human users vs. legitimate social bots and human spammers vs. social bot spammers. Furthermore, the authors suggest studying the impact of increased tweet length on automated spam accounts' ability to generate intelligent content due to recent changes in maximum tweet length limits. Overall, this spam account detection approach can be applied not only to Twitter but also to other social media platforms or real-time filtering applications for data collection pipelines. The study provides valuable insights and lays the foundation for future research in this field.

Was this synthesis useful?

Score: 0

Why do we need votes?

Votes are used to determine whether we need to re-run our summarizing tools. If the count reaches -10, our tools can be restarted.

Figures and plots

Ask about a figure by number or use its question button to request visual analysis. Up to two figures and four image panels per question. Values read from plots are approximate; check the original data.

AI assistant instructions?

Tools for this paper

Explain it differently 1 free rewrite left today
Save to a collection Group papers in collections to compare them, chat with them and export them. Continue with Google Create a free account Log in

Advanced tools Premium

For researchers who follow a field closely: citation alerts, journal-club slides, version comparison and Zotero export. Discover Premium

Zotero Zotero guide
Citation alerts 9 papers already cite it: be emailed each new one. Citation alerts guide
Who works on it?

Among the 31 papers around this one, the name that comes back most: Vasilii A. Gromov (2 papers)

1 more person signs several of them. The full map adds their institutions and approaches, from the papers of the last years.

Continue with Google Create a free account: your first map is free
Weekly letter The latest letter of its field, "Artificial intelligence this week" of 16 November 2026: Artificial intelligence this week: What impressive scores leave out · Receive one by email for free

Papers around this one

See them on a map

Ranked by how close their title and abstract are to this paper’s. Select some to read them together.

Closest papers

  1. Very close Another approach to the problem
    Social media spammer identification using deep learning
    C. S. Santhosh Kumar, A. Kiran, Vinay Kumar et al. · 2024 · MATEC Web of Conferences · cited 1 time · Not summarized yet

    Leveraging Machine Learning Techniques and NLP detects fake social-network accounts using metadata and text, sharing this paper’s fake-account detection task and NLP component.

    Abstract

    Social media has become a part of our lives with countless individuals actively sharing content and connecting with others. However this open platform also presents a challenge, in dealing with misinfor- mation and the spread of information by individuals. Social media platforms face the task of identifying these users and preventing the dissemination of content. In this study the focus is on Twitter as a case example. Explore the use of deep learning techniques to detect fake accounts. We analyze factors such as account age, follower to following ratio, tweet frequency and URL to tweet ratio. Through comparing algorithms the aim is to determine the effective approach, for identifying and co…

  2. Very close Same question
    Leveraging Machine Learning Techniques and NLP for Identifying Fake Accounts in Social Networks
    V. Siddhartha · 2025 · International Journal for Research in Applied Science and Engineering Technology · Not summarized yet

    This project detects fake social-network accounts using profile metadata and NLP, while the paper detects fake Twitter users using NLP and graph embeddings.

    Abstract

    The rise of malicious entities on social networking platforms has led to a significant increase in the creation and use of fake accounts, which can be exploited for misinformation, cybercrime, and social manipulation. To address this challenge, this project presents a machine learning-based approach for the detection of fake accounts using both structured metadata and natural language processing techniques. This project utilizes a comprehensive feature extraction process, considering username patterns, profile statistics, and textual metadata, followed by preprocessing and normalization. A comparative analysis of multiple classifiers - Random Forest, Support Vector Machine (SVM), Logistic Re…

  3. Very close Same question
    A Feature Based Approach to Detect Fake Profiles in Twitter
    Jyoti Kaubiyal, A. Jain · 2019 · International Conference on Big Data and Internet of Things · cited 27 times · Not summarized yet

    A Feature Based Approach detects fake profiles on Twitter, the same platform and account-identification problem studied here.

    Abstract

    Social networking platforms, particularly sites like Twitter and Facebook have grown tremendously in the past decade and has solicited the interest of millions of users. They have become a preferred means of communication, due to which it has also attracted the interest of various malicious entities such as spammers. The growing number of users on social media has also created the problem of fake accounts. These false and fake identities are intensively involved in malicious activities such as spreading abuse, misinformation, spamming and artificially inflating the number of users in an application to promote and sway public opinion. Detecting these fake identities, thus becomes important to…

  4. Very close Same question
    Spammer Detection And Fake User Identification On Social Networks
    Aditya Verma, Devansh Badoni, Rahul Chauhan et al. · 2023 · 2023 IEEE Technology & Engineering Management Conference - Asia Pacific (TEMSCON-ASPAC) · Not summarized yet

    Spammer Detection And Fake User Identification proposes supervised learning to identify Twitter spammers, addressing the same fake-user detection problem.

    Abstract

    Spammers who seek to disseminate tonnes of dangerous and useless material have turned popular social networking sites into a target. For instance, Twitter has become one of the most frequently utilised platforms ever and enables an excessive amount of spam as a result. The detection of spammers and the identifying of fake Twitter users is a major area of research in today’s online social networks. In this paper, we examine the techniques for identifying Twitter spammers. A supervised machine learning-based strategy for spammer identification is proposed in this study. Gathering a dataset from Twitter that contains user messages is the first and most important stage in the procedure. to separ…

  5. Very close Same question Cited by this paper Used in depth by this paper
    Detection of spam-posting accounts on Twitter
    Isa Inuwa-Dutse, Mark Liptrott, Ioannis Korkontzelos · 2018 · Neurocomputing · cited 140 times · Not summarized yet

    Detection of spam-posting accounts identifies spam or fake users on social media, the target this paper detects on Twitter.

    Abstract

    Abstract Online Social Media platforms, such as Facebook and Twitter, enable all users, independently of their characteristics, to freely generate and consume huge amounts of data. While this data is being exploited by individuals and organisations to gain competitive advantage, a substantial amount of data is being generated by spam or fake users. One in every 200 social media messages and one in every 21 tweets is estimated to be spam. The rapid growth in the volume of global spam is expected to compromise research works that use social media data, thereby questioning data credibility. Motivated by the need to identify and filter out spam contents in social media data, this study presents …

  6. Very close Same question
    Spam Detection and Fake User Identification in Twitter: An Analysis of Machine Learning Models
    R. S, Phani Sri Siva Charan K, S. M et al. · 2023 · International Journal of Advanced Research in Science, Communication and Technology · Not summarized yet

    Spam Detection and Fake User Identification compares machine-learning methods for Twitter spammer and fake-user detection, the same platform and task.

    Abstract

    Spammer detection and fake user identification are significant issues in the realm of social media, with Twitter being no exception. The detection of spammers and fake users on Twitter is essential for preserving the platform's integrity and protecting users from online scams and fraud. This project paper aims to conduct a comprehensive study of the different techniques and algorithms used for spammer detection and fake user identification on Twitter. We will evaluate the effectiveness of traditional techniques and machine learning-based methods and propose a novel approach combining both methods for spammer detection and fake user identification on Twitterpils.

  7. Very close Same question
    Spammer Detection and Fake User Identification on Social Networks
    Faiza Masood, Ghana Ammad, Ahmad S. Almogren et al. · 2019 · IEEE Access · cited 109 times · Not summarized yet

    Spammer Detection and Fake User Identification studies spam and fake users on social networks, including Twitter, matching this paper’s detection target.

    Abstract

    Social networking sites engage millions of users around the world. The users’ interactions with these social sites, such as Twitter and Facebook have a tremendous impact and occasionally undesirable repercussions for daily life. The prominent social networking sites have turned into a target platform for the spammers to disperse a huge amount of irrelevant and deleterious information. Twitter, for example, has become one of the most extravagantly used platforms of all times and therefore allows an unreasonable amount of spam. Fake users send undesired tweets to users to promote services or websites that not only affect legitimate users but also disrupt resource consumption. Moreover, the pos…

  8. Very close Another approach to the problem
    Fake Profile Identification in Social Network using Machine Learning and NLP
    L. P, S. V., V. Sasikala et al. · 2022 · 2022 International Conference on Communication, Computing and Internet of Things (IC3IoT) · cited 19 times · Not summarized yet

    Fake Profile Identification uses machine learning and NLP to classify genuine and fake social-network accounts, while this paper adds graph embeddings for Twitter.

    Abstract

    In present times, social media plays a key role in every individual life. Everyday majority of the people are spending their time on social media platforms. The number of accounts in these social networking sites has dramatically increasingday-by-day and many of the users are interacting with others irrespective of their timeand location. These social media sites have both pros and cons and providesecurity problems to us also for our information. To scrutinize, who are giving threats in thesenetworking sites we need to organize these social networking accounts into genuine accountsand fake accounts. Traditionally, we are having different classification methods to point outthe fake accounts o…

Prior work it builds on

  1. Close Same question A method it uses
    Detecting Fake Accounts on Social Media
    S. Khaled, N. El-Tazi, Hoda M. O. Mokhtar · 2018 · 2018 IEEE International Conference on Big Data (Big Data) · cited 117 times · Not summarized yet

    Detecting Fake Accounts on Social Media investigates account features and classification algorithms to identify fake accounts, the same core task.

    Abstract

    In the present generation, on-Line social networks (OSNs) have become increasingly popular, people’s social lives have become more associated with these sites. They use on-Line social networks (OSNs) to keep in touch with each others, share news, organize events, and even run their own e-business. The rapid growth of OSNs and the massive amount of personal data of its subscribers have attracted attackers, and imposters to steal personal data, share false news, and spread malicious activities. On the other hand, researchers have started to investigate efficient techniques to detect abnormal activities and fake accounts relying on accounts features, and classification algorithms. However, some…

  2. Close Broader context A method it uses
    An in-depth characterisation of Bots and Humans on Twitter
    Zafar Gilani, R. Farahbakhsh, Gareth Tyson et al. · 2017 · arXiv.org · preprint, not peer reviewed · cited 29 times · Not summarized yet

    An in-depth characterisation compares bots and humans on Twitter using tweet metadata and annotated labels, examining account types related to this paper’s targets.

    Abstract

    Recent research has shown a substantial active presence of bots in online social networks (OSNs). In this paper we utilise our past work on studying bots (Stweeler) to comparatively analyse the usage and impact of bots and humans on Twitter, one of the largest OSNs in the world. We collect a large-scale Twitter dataset and define various metrics based on tweet metadata. We divide and filter the dataset in four popularity groups in terms of number of followers. Using a human annotation task we assign 'bot' and 'human' ground-truth labels to the dataset, and compare the annotations against an online bot detection tool for evaluation. We then ask a series of questions to discern important behav…

  3. Close Another approach to the problem
    Online Human-Bot Interactions: Detection, Estimation, and Characterization
    Onur Varol, Emilio Ferrara, Clayton A. Davis et al. · 2017 · International Conference on Web and Social Media · cited 1,157 times · Not summarized yet

    Online Human-Bot Interactions detects Twitter bots using user, content, network, and activity features, an alternative approach to this paper’s fake-user detection.

    Abstract

    Increasing evidence suggests that a growing amount of social media content is generated by autonomous entities known as social bots. In this work we present a framework to detect such entities on Twitter. We leverage more than a thousand features extracted from public data and meta-data about users: friends, tweet content and sentiment, network patterns, and activity time series. We benchmark the classification framework by using a publicly available dataset of Twitter bots. This training data is enriched by a manually annotated collection of active Twitter users that include both humans and bots of varying sophistication. Our models yield high accuracy and agreement with each other and can …

  4. Related* Another approach to the problem A method it uses
    Bot Detection on Online Social Networks Using Deep Forest
    Kheir Eddine Daouadi, R. Rebaï, I. Amous · 2019 · Computer Science On-line Conference · cited 29 times · Not summarized yet

    Bot Detection on Online Social Networks uses Deep Forest to detect bots, addressing a related account-detection problem with a different model.

  5. Related Broader context
    BotCamp: Bot-driven Interactions in Social Campaigns
    Noor Abu-El-Rub, Abdullah Mueen · 2019 · The Web Conference · cited 34 times · Not summarized yet

    BotCamp identifies political bots and campaigns through temporal, textual, and network clustering, focusing on coordinated activity rather than individual fake-user classification.

    Abstract

    Bots (i.e. automated accounts) involve in social campaigns typically for two obvious reasons: to inorganically sway public opinion and to build social capital exploiting the organic popularity of social campaigns. In the process, bots interact with each other and engage in human activities (e.g. likes, retweets, and following). In this work, we detect a large number of bots interested in politics. We perform multi-aspect (i.e. temporal, textual, and topographical) clustering of bots, and ensemble the clusters to identify campaigns of bots. We observe similarity among the bots in a campaign in various aspects such as temporal correlation, sentimental alignment, and topical grouping. However, …

Papers that cite it

  1. Close Builds on this paper
    ATTENTION-DRIVEN TABULAR DEEP LEARNING FRAMEWORK FOR DETECTING FAKE PROFILES IN SOCIAL NETWORKING PLATFORMS
    Suba Suba, S. Geetha, S. Prabakaran et al. · 2026 · International Journal of Drug Delivery Technology · Not summarized yet

    Attention-Driven Tabular Deep Learning cites this paper and presents an attention-based model for detecting fake profiles on social platforms.

    Abstract

    Background The quick development of the social networking sites has brought about more opportunities in communication and exchange of information, but it has also enhanced the existence of fake user accounts that indulge in activities like spamming, spreading misinformation and fraud over the internet. Such bad actors can pose a massive credibility and security risk to social media ecosystems. As such, the issue of coming up with effective and scalable systems of detecting the fraud profiles has now been a significant research question. Objective The paper describes a deep learning model of attention-based tabular detection of fake profiles in social networking platforms. Materials and Metho…

  2. Close Same method, another problem
    Machine Learning Boosting Algorithms to Detect the Fake User Profiles on Instagram
    D. Naik, S. Sri, G. Priya et al. · 2023 · 2023 International Conference on Quantum Technologies, Communications, Computing, Hardware and Embedded Systems Security · cited 4 times · Not summarized yet

    Machine Learning Boosting Algorithms applies boosting methods to detect fake Instagram profiles, using machine learning for a different platform.

    Abstract

    Social media use has become a significant part of every person’s daily life during the past several years. These social media sites support in the three Cs: creating, connecting, and communicating with others. Nearly 4.8 billion people use social media globally, according to use data. There are millions of individuals worldwide that spend the majority of their time on social networking websites. A variety of threats, such as bogus profiles, account hacking, safety and security problems, might arise with an increase in usage. The most popular and broadly used social networking site is Instagram. The purpose of the following paper is to identify phoney accounts of people on Instagram using mac…

  3. Close Builds on this paper
    DeepMUI: A novel method to identify malicious users on online social network platforms
    Akash Shah, S. Varshney, Monica Mehrotra · 2023 · Concurrency and Computation · cited 13 times · Not summarized yet

    DeepMUI cites this paper and develops a method for identifying malicious users on online social networks.

    Abstract

    The use of online social network (OSN) platforms has become an essential component of contemporary society, facilitating global connectivity, and information sharing among individuals. The proliferation of malicious users has emerged as a noteworthy obstacle, exerting a detrimental effect on the authenticity of the data disseminated through these channels. A malicious profile is created with the intention of disseminating false information, manipulating perspectives, and executing harmful actions, including phishing schemes, identity theft, and the propagation of malware. Consequently, the identification of malicious users has emerged as an essential undertaking for both OSN platforms and re…

  4. Close Review
    Evaluating Machine Learning Strategies For The Detection Of Fake Profiles: A Survey
    Atithi Negi, Bhoomi Sharma, Shubham Kumar · 2025 · 2025 International Conference on Emerging Technologies and Innovation for Sustainability (EmergIN) · Not summarized yet

    Evaluating Machine Learning Strategies reviews machine-learning approaches and challenges for fake-profile detection across social media.

    Abstract

    The swift development of social media has transformed various domains like business, communication and the way public discussions take place globally. But along with the benefits comes the downside as it has led to a rapid growth in deceptive profiles which are used for evil motives like harming an individual or an organization, spreading false information, impersonation etc. This weakens trust and authenticity of an online space and creates risk for both users and digital platforms. In this paper, we review the challenges and solutions associated with false profile detection across social media platforms. It highlights the pivotal role of Machine Learning (ML) in automating and improving th…

  5. Close* Same method, another problem
    A textual fingerprint learning model to detect fake information spreaders in social networks
    Rashid Behzadidoost, Habib Izadkhah · 2025 · Neurocomputing · Not summarized yet

    A textual fingerprint learning model uses text-based learning to identify fake-information spreaders in social networks, applying a related textual approach to another target.

How are these papers found?

We compare the title and abstract of this paper with those of every paper summarized on this site, and with the papers Semantic Scholar recommends, those it cites and those citing it (for a much-cited paper, its most recent citations and, from OpenAlex, the most cited ones). Among the prior work and the follow-ups, the papers that use each other in depth, a method it uses and the much-cited follow-ups come a little higher. The levels (very close, close, related) come from the similarity of the texts, not from a judgment of quality; * marks a paper without an abstract, compared by its title. The sentence under a paper is written by an AI from the two abstracts: check it in the papers. A paper marked not relevant by three readers leaves the list.

Updated on 3 November 2026.

The illustrated guide to the papers around a paper

The previous summary was created more than a year ago and can be re-run (if necessary) by clicking on the Run button below.

Disclaimer: The AI-based summarization tool and virtual assistant provided on this website may not always provide accurate and complete summaries or responses. We encourage you to carefully review and evaluate the generated content to ensure its quality and relevance to your needs.