The Spread Of Misinformation In Online Social Networks

In recent years,with the rapid developing of the Internet,Online Social Networks(OSN)have changed the way people produce and consume information.The prosperity of OSNs has greatly improved the quality of life for people.However,massive amounts of misinformation have been spreading over social media,causing wide public concern about our information ecosystem.Now it has been identified as a major global risk.How does fact checking compete with misinformation for user attentions? How do misinformation and fact checking differ in their spreading patterns? And what strategies are used by social bots to promote the spread of misinformation? While researchers from communication,cognitive,social,and computer are studying the complex causes for the viral diffusion of misinformation,little systematic,data-based evidence has been published to guide these efforts.This thesis will make a systematic and quantitative study on the spread of misinformation through a big data mining view.The main contributions of this thesis are as follows.(1)We designed and implemented a source-based system to collect data on the spread of misinformation and fact checking from the public Twitter stream and news websites.When studying misinformation,the first challenge is to assess the truthfulness of a claim:it is impossible to manually evaluate a very large number of claims,even for professional fact-checking organizations with dedicated staff.Here we mitigate these issues by relying on a list of low-credibility sources compiled by trusted third-party organizations.Compared with other systems,our system is proved to be scalable and effective.The dataset,consisting of about 1 million articles and 50 millions of tweets,is one of the largest datasets in the current study of misinformation.(2)We presented the anatomy of an online misinformation network.Thanks to the data captured by Hoaxy,we are able to study the misinformation diffusion network.We perform k-core decomposition on a diffusion network obtained from two million retweets produced by several hundred thousand accounts over the six months before the election.As we move from the periphery to the core of the network,fact-checking nearly disappears,while social bots proliferate.The number of users in the main core reaches equilibrium around the time of the election,with limited churn and increasingly dense connections.We conclude by quantifying how effectively the network can be disrupted by penalizing the most central nodes.These findings provide a first look at the anatomy of a massive online misinformation diffusion network.(3)We characterized the competition of misinformation and fact checking.Started with Snopes(snopes.com),we conducted(a)a survival analysis that shows that about70% of claims will be fact-checked in one week and about 800 tweets with claim links are posted during this period,on average;(b)a distribution of article popularity that shows that false claims can become more popular than the corresponding debunking;(c)a distribution of types of tweets that shows that fact checking tends to spread in a conversational way.Then we replicated some analysis on a much larger database in a statistical way.The conclusions of the statistical analysis are consistent with those of the case study.In addition,we focused on abnormal patterns held by the spread of misinformation,including(a)a single account can post the same claim article even thousands of times;(b)the more a story was tweeted,the more the tweets were concentrated in the hands of few accounts,who act as ”super-spreaders”;(c)the super-spreaders are significantly more likely to be bots compared to the general population of accounts who share misinformation.The findings concluded that social bots could play an important role in the spread of misinformation.(4)We studied the role played by social bots in the spread of misinformation.Started by cases,we explored the strategies used by social bots.Then we study it in a statistical way and find evidence that social bots played a disproportionate role in spreading of misinformation.Bots amplify misinformation content in the early spreading moments,before an article goes viral.They also target users with many followers through replies and mentions.Humans are vulnerable to this manipulation,resharing content posted by bots.Successful low-credibility sources are heavily supported by social bots.These results suggest that curbing social bots may be an effective strategy for mitigating the spread of online misinformation.