DeepScan: Exploiting Deep Learning for Malicious Account Detection in Location-Based Social Networks

Gong, Qingyuan; Chen, Yang; He, Xinlei; Zhou, Zhuang; Wang, Tianyi; Huang, Hong; Wang, Xin; Fu, Xiaoming

doi:10.1109/mcom.2018.1700575

Cited by 72 publications

(32 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the future, we plan to analyze the Airbnb users' online behavior and offline activities as an integrated whole. Also, we aim to detect the spam accounts using deep learning technologies [10].…”

Section: Discussionmentioning

confidence: 99%

Measurement and Analysis of the Reviews in Airbnb

Zhou¹,

Chen²,

Ma³

et al. 2018

2018 IFIP Networking Conference (IFIP Networking) and Workshops

Self Cite

View full text Add to dashboard Cite

Airbnb, a recently emerged online lodging service that allows house and apartment dwellers to lease out their premises to short-term renters like tourists, is reconstructing the value chain of the traditional hotel industry. It works as a platform that connects hosts and travelers and facilities their interaction and exchange. Studying this service could shed light on understanding the emerging sharing economy from a usercentric perspective. In this work, we collect the profiles of 43.8 million Airbnb users, and analyze the reviews they published online. We model the interactions between Airbnb users using a review graph, and study their mobility patterns by investigating their reviews. To the best of our knowledge, our work is the first measurement study of massive Airbnb users on a global scale, and it provides insights of their activities in both cyberspace and the physical world.

show abstract

Section: Discussionmentioning

confidence: 99%

Measurement and Analysis of the Reviews in Airbnb

Zhou¹,

Chen²,

Ma³

et al. 2018

2018 IFIP Networking Conference (IFIP Networking) and Workshops

Self Cite

View full text Add to dashboard Cite

show abstract

“…The current study discusses fake account detection in social networks by using the learning method. DeepScan proposes malicious account detection with DL in Location-Based Social Networks (LBSN) [23]. Another scheme utilizes the similarity of the user's friends to calculate the adjacency matrix of the graph to identify the user as benign or fake [24].…”

Section: Osn Link Featuresmentioning

confidence: 99%

RunPool: A Dynamic Pooling Layer for Convolution Neural Network

Huang¹,

Wanda²

2020

IJCIS

View full text Add to dashboard Cite

A B S T R A C TDeep learning (DL) has achieved a significant performance in computer vision problems, mainly in automatic feature extraction and representation. However, it is not easy to determine the best pooling method in a different case study. For instance, experts can implement the best types of pooling in image processing cases, which might not be optimal for various tasks. Thus, it is required to keep in line with the philosophy of DL. In dynamic neural network architecture, it is not practically possible to find a proper pooling technique for the layers. It is the primary reason why various pooling cannot be applied in the dynamic and multidimensional dataset. To deal with the limitations, it needs to construct an optimal pooling method as a better option than max pooling and average pooling. Therefore, we introduce a dynamic pooling layer called RunPool to train the convolutional neural network (CNN) architecture. RunPool pooling is proposed to regularize the neural network that replaces the deterministic pooling functions. In the final section, we test the proposed pooling layer to address classification problems with online social network (OSN) dataset.

show abstract

“…Deep convolutional neural networks have been also investigated to analyze sentiments in Twitter [28]. Deep learning based methods have been used to detect malicious accounts in location-based social networks [29]. One recent work used a Bayesian network and fuzzy recurrent neural networks for detecting subjectivity [30].…”

Section: Related Workmentioning

confidence: 99%

Characterization of citizens using word2vec and latent topic analysis in a large set of tweets

Vargas-Calderón

Camargo

2019

Cities

View full text Add to dashboard Cite

With the increasing use of the Internet and mobile devices, social networks are becoming the most used media to communicate citizens' ideas and thoughts.This information is very useful to identify communities with common ideas based on what they publish in the network. This paper presents a method to automatically detect city communities based on machine learning techniques applied to a set of tweets from Bogotá's citizens. An analysis was performed in a collection of 2,634,176 tweets gathered from Twitter in a period of six months.Results show that the proposed method is an interesting tool to characterize a city population based on a machine learning methods and text analytics. reducing the digital gap in Colombia. For instance, from 2011 to 2014, several portions of the city have experienced an increase of up to 148.3% in homes with Internet connection [2]. By 2011, both Facebook and Twitter were the most popular social networks in Bogotá [1]. Moreover, 65.5% of Bogotá's Internet users used social networks by 2014 [2]. This study concentrates on Twitter because it allows its users to write posts with a length ranging from 1 to 140 characters 1 , making Twitter a tool for microblogging, a form of communication in which users express their opinion about several topics in short posts (tweets) [3].By only taking the users' posts and not the explicit relations between the users, our large data set is an interesting basis for testing unsupervised community detection methodologies. In fact, the only common factor that relates the users from our data set is that they are connected by geographical location.Such unsupervised community detection methodologies help to understand social phenomena that takes place in that geographical region and in a particular period of using data from any social network with no a priori knowledge of people's relations.With the objective of identifying the main topics treated by Bogotá's population on Twitter, as well as detecting possible communities, we collected tweets emitted from Bogotá. Then, a Word2Vec model [4] was built in order to represent the set of tweets corresponding to each user as a vector in a vector space.The gap statistic [5] was used to estimate the number of clusters that could be formed using these vectors. Finally, a frequency distribution of words was built for each cluster, so that each cluster could be identified by its most frequent words, which ultimately characterizes a topic.Our work acknowledges that online social networks constitute one of the main scenarios where people express their opinions, which makes them an outstanding source of information that allows the characterization of important topics for the citizenship. Therefore, we provide a robust method for studying 1 Data was collected before the change in the maximum tweet length to 280 characters

show abstract

DeepScan: Exploiting Deep Learning for Malicious Account Detection in Location-Based Social Networks

Cited by 72 publications

References 9 publications

Measurement and Analysis of the Reviews in Airbnb

Measurement and Analysis of the Reviews in Airbnb

RunPool: A Dynamic Pooling Layer for Convolution Neural Network

Characterization of citizens using word2vec and latent topic analysis in a large set of tweets

Contact Info

Product

Resources

About