Real-Time Bot Detection from Twitter Using the Twitterbot+ Framework

Daouadi, Kheir Eddine; Rebaï, Rim Zghal; Amous, Ikram

doi:10.3897/jucs.2020.026

Cited by 7 publications

(3 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They proposed a scalable and generalizable bot detection method and used a data selection criterion to find the best model. Most of the methods that use only the metadata information from the user profile are trained using Random Forest or Adaboost classifiers (Daouadi et al 2020 ; Kondeti et al 2021 ). Deep learning techniques are not much explored when using these sets of features.…”

Section: Related Workmentioning

confidence: 99%

DeeProBot: a hybrid deep neural network model for social bot detection based on user profile data

Hayawi

Mathew

Venugopal

et al. 2022

Soc. Netw. Anal. Min.

View full text Add to dashboard Cite

Use of online social networks (OSNs) undoubtedly brings the world closer. OSNs like Twitter provide a space for expressing one’s opinions in a public platform. This great potential is misused by the creation of bot accounts, which spread fake news and manipulate opinions. Hence, distinguishing genuine human accounts from bot accounts has become a pressing issue for researchers. In this paper, we propose a framework based on deep learning to classify Twitter accounts as either ‘human’ or ‘bot.’ We use the information from user profile metadata of the Twitter account like description, follower count and tweet count. We name the framework ‘DeeProBot,’ which stands for Deep Profile-based Bot detection framework. The raw text from the description field of the Twitter account is also considered a feature for training the model by embedding the raw text using pre-trained Global Vectors (GLoVe) for word representation. Using only the user profile-based features considerably reduces the feature engineering overhead compared with that of user timeline-based features like user tweets and retweets. DeeProBot handles mixed types of features including numerical, binary, and text data, making the model hybrid. The network is designed with long short-term memory (LSTM) units and dense layers to accept and process the mixed input types. The proposed model is evaluated on a collection of publicly available labeled datasets. We have designed the model to make it generalizable across different datasets. The model is evaluated using two ways: testing on a hold-out set of the same dataset; and training with one dataset and testing with a different dataset. With these experiments, the proposed model achieved AUC as high as 0.97 with a selected set of features.

show abstract

Section: Related Workmentioning

confidence: 99%

DeeProBot: a hybrid deep neural network model for social bot detection based on user profile data

Hayawi

Mathew

Venugopal

et al. 2022

Soc. Netw. Anal. Min.

View full text Add to dashboard Cite

show abstract

“…Technology advancement has resulted in the emergence of many new generation digital media and communication platforms today. Social media has become an integral component of daily life since billions of users worldwide including social media researchers, companies, politicians, and even government agencies come together in a virtual environment every day on these global platforms [1,2]. Today, most people use social media for interaction, entertainment, socialization, information seeking and sharing, self-expression, education and surveillance [3].…”

Section: Introductionmentioning

confidence: 99%

Automatic Detection of Disinformation: A Systematic Mapping Study

ÖNCÜL,

İRKEY,

GÖK

et al. 2023

Journal of Polytechnic

View full text Add to dashboard Cite

In recent years, the spread of disinformation, which is one of the kind of information pollution, has accelerated on online social media platforms, and detecting disinformation early has become significant to be able to remove the negative impact it has on individuals and societies. In this direction, increased number of studies focusing on the automatic detection of disinformation and the variety of approaches developed have been observed in recent years, and the need to study the trends in the studies carried out in detail has emerged. This research seeks to present a map of the research area for the automatic detection of disinformation. In this context, 61 primary sources published in the electronic databases named Web of Science and Scopus between 2018-2022 included in the research scope have been examined and analyzed within the framework of the determined criteria. The conducted systematic mapping study aims to provide useful insights about automatic detection of disinformation including publication year, journal, journal class and publisher name, country of origin of the authors, most prolific authors and institutions, keywords used, supported approaches, obtained accuracy rates and datasets used. It is expected that this research will guide/direct researchers about the approaches developed for the detection of disinformation and contribute to future studies.

show abstract

“…With a lot of personal views about a product posted online, consumer reviews often play an effect on potential buyer behaviour when it comes to online purchasing [52]. There has been a lot of research performed in various fields such as identifying rumors in Twitter [44], identifying bots in Twitter [16] and bias detection in politics [6] given it's public access to data. Twitter is one such social media site that is of major focus by companies to perform research on their target audience and understand the opinion of the masses.…”

Section: Introductionmentioning

confidence: 99%

Sentiment Analysis Framework using Deep Active Learning for Smartphone Aspect Based Rating Prediction

Muralidhar

Hulipalled

2023

Foundations of Computing and Decision Sciences

View full text Add to dashboard Cite

Social media are a rich source of user generated content where people express their views towards the products and services they encounter. However, sentiment analysis using machine learning models are not easy to implement in a time and cost effective manner due to the requirement of expert human annotators to label the training data. The proposed approach uses a novel method to remove the neutral statements using a combination of lexicon based approach and human effort. This is followed by using a deep active learning model to perform sentiment analysis to reduce annotation efforts. It is compared with the baseline approach representing the neutral tweets also as a part of the data. Considering brands require aspect based ratings towards their products or services, the proposed approach also categorizes predicting ratings of each aspect of mobile device.

show abstract

Real-Time Bot Detection from Twitter Using the Twitterbot+ Framework

Cited by 7 publications

References 2 publications

DeeProBot: a hybrid deep neural network model for social bot detection based on user profile data

DeeProBot: a hybrid deep neural network model for social bot detection based on user profile data

Automatic Detection of Disinformation: A Systematic Mapping Study

Sentiment Analysis Framework using Deep Active Learning for Smartphone Aspect Based Rating Prediction

Contact Info

Product

Resources

About