From university desks to EIA Regional Winners: the journey of team Bot Busters
Most of the time, people choose to pursue a Ph.D. because of their curiosity. This attitude leads them to be always hungry for knowledge, trying to audit every possible source of information.
In our case, we read about the competition through a minor university channel: the usual mailing list I’d say, where thousands of people share information knowing that probably no one will ever read it. However, we are Ph.D. students, do care about reading information. In this way, we got in touch with the EIA 2018: a students’ competition sponsored by a leader telco company Ericsson, really nice. The topic of this year is The Future of Truth: it means to explore the ways in which it is possible to collect, share and validate information, especially in this modern era, interesting. Watching the promotional video, some keywords caught my attention: “cyber security”, “artificial intelligence”, “data mining.” My mind stuck on that light-blue panel, and I thought immediately to share this news with my workmates. “Why don’t we submit a proposal? We have the necessary expertise in this field, and we have the chance to be the ones leading a real proposal!”. No need to be convinced, everyone was already in.
That’s how we built our international team, which features a strong partnership between the universities of Murcia, in Spain, and of the Aegean, in Greece. Pantaleone, Mattia and Dimitrios are three Ph.D. students with experience in solving cybersecurity challenges. This time, we’re working together for validating and sharing the truth in an era when the credibility of the world-wide-web becomes a controversial matter.
My name is Mattia, I’m an Italian 2nd year Ph.D. candidate at the University of Murcia and I’m a certified Italian Engineer that researches the implications of modelling user behaviours in the cybersecurity. Dimitrios, our Greek member, is a 4th year Ph.D. candidate at the University of the Aegean, with knowledge in the application of artificial intelligence and machine learning to detect cyber intrusions. Last, but not least, Pantaleone, also Italian, is a 1st year Ph.D. student at the University of Murcia, specialised in artificial intelligence approaches to react against cyber-attacks.
Bot Busters: Our Proposal
While shaping the proposal, our idea was to leverage our expertise in the cybersecurity ecosystem to tackle a real problem in the modern society, something with a real impact on the people. Following this reasoning, we started to build our project, with a particular focus on how to fight the “bad guys” on the most used platforms nowadays, which is social media.
Undoubtedly, social networks have changed our lives. For many, the number of followers and in general social interactions are a real-world asset, and not something relegated to the digital world. Moreover, according to the last reports, over two-thirds of US residents are actively interacting with news in social networks instead of classic mass media. Social networks are a constellation of interconnected profiles, although some of them, with much more connections than the others, are acting as information hubs. Those profiles are commonly indicated as influencers.
The History of Social Bots
The idea of leveraging influencers for marketing purposes can be traced back up to the mid-1990s when the idea of forging the public opinion under the influence of the so-called, opinion leaders took ground. Users then became opinion followers, and the social influence became a peculiarity of specific elite groups. Typically, opinion leaders are held in high esteem by those who accept their opinions, and their endorsements are interpreted as highly credible electronic word-of-mouth rather than paid advertising. This indistinguishability between paid content and the organic information is at the very base of any public opinion manipulation approach. Creating influencers using forged accounts is not only possible, but it has been done successfully in the past few years – sometimes with dramatic consequences in the public opinion.
Recently, it has been proved by academic researchers how political actors are also leveraging the potentiality of social botnets to sneakily manipulate public opinion online. Political parties are, in fact, emulating the tactic of purchasing massive amounts of followers to significantly increase the spread of the propaganda and amplify the extension of political spam: media reports have been uncovered the usage of social bots for political purposes across the globe, from Australia to the United Kingdom and from Venezuela to Russia.
The Social Bots spread and usage is without questions a large and systematic problem athwart all the social networks. Social bots suppress free expression and public innovation through the demobilization and the suffocation of democratic free speech. They subtle manipulate public opinion by giving false impressions of influencer popularity. To give a proof of their powerful impact, it has been proved that they had actively taken part in promoting ISIS propaganda, spreading rumours and conspiracy theories, and manipulating the stock market. Unfortunately, to face these problems both the industrial and the academic worlds needs to put an extra effort in researching and developing technologies to increase the awareness of the final users as an ultimate goal. Recently, also the New York times started considering this problem, publishing an inquiry that caused the dismantle of a botnet featuring millions of forged accounts. Fortunately, the Ericsson Innovation Award 2018 gives us the opportunity to rise this threat to a higher level, impacting more and more people. We leveraged this chance to present our ideas, proposing a solution which may concretely help people in this direction.
Bot Busters Social Detection Framework
Following this reasoning, we presented our social detection framework: Bot Busters; it rises to shield users against crafted influencers and forged networks of social bots. Together, we will take advantage of artificial intelligence to tackle the social botnets, aiming to increase the awareness of the average-user alongside with providing insight tools to experts and marketing strategists. Bot Busters will construct an automated framework for ultimately increasing the awareness toward social botnets’ usage and strategies, with special attention and dedication to exposing the accountability of the main influencers.
The team acknowledges and thanks the support provided by all the Ericsson Mentors during these seven weeks. It is a great opportunity for us to exchange ideas with highly qualified people and witness how ideas are transformed into valuable products.
Follow the Bot Buster team on the official webpage (www.socialbotbusters.com) and the social networks. Stay tuned for news, demo and proof-of-concepts.