Robert Ackland: Hyperlinks are connecting pages together, and allowing people to, as they surf the web, find new information. This, for me, has always been the thing that I’ve been most interested in, because there is a social science of why hyperlinks are created. And what does it mean for a website to create a hyperlink to another website? It’s used in order to guide people’s attention, shape people’s attention. And so the types of actors that I studied, have been political parties, social movements, organizations, activists, and they are all making choices about how to hyperlink to, and why. And these choices have measurable impacts on shaping the attention of other people.
Noshir Contractor: My guest today is Professor Robert Ackland from the School of Sociology at the Australian National University in Canberra. You just heard him talk about his work with hyperlinks. Rob works at the intersection of network science and web science, to study networks on the Web. Under a 2005 special initiative of the Australian Research Council, he established the Virtual Observatory for the Study of Online Networks, VOSON for short. His research has been published in journals such as Social Networks, Journal of Social Structure, Computational Economics and Social Science Computer Review.And his book, Web Social Science: Concepts, Data and Tools for Social Scientists in the Digital Age was published in 2013.
Robert Ackland: Great to be here, Nosh.
Noshir Contractor: Welcome, Rob. Thanks again for joining us from downunder. I want to start by asking you what got you interested in web science, coming as you did from an economics background, initially interested in issues that were related to economic development?
Robert Ackland: As you mentioned, Nosh, all my training is actually in economics. In my first academic job after my PhD, in 2001, I was working in an interdisciplinary research center, and I started working with a political scientist, Rachel Gibson, who’s now at the University of Manchester. And Rachel and colleagues were working on studying how political parties were using the web in order to undertake various political functions, such as raising awareness about issues, engaging with potential voters raising revenue. A big aspect of that work revolved around the hyperlink — the idea that if you have more hyperlinks pointing to your website, that can bring more eyeballs to your content, and therefore allow you to raise more awareness about issues that could concern you.
I saw an opportunity there to use web crawlers to collect large scale hyperlink network data sets, and then to start studying these networks. I produced, for example, networks of political parties, looking at mainstream versus major parties, and conservative versus, liberal parties. So I started to look at the structure of hyperlinks of political parties.
I realized that what I was doing was, in fact, social network analysis applied to the web. And a big part of my career has been looking at how can methods and approaches from social network analysis be adapted to study online networks?
Noshir Contractor: You use the word hyperlink a few times. And I know that that word just rolls off your tongue with a lot of ease. But for most people, when we think of social networks, we think of potentially, links between people. So you have a friend on a social media platform or a follower on a social media platform. But when you’re talking about hyperlinks, these are links not between people, but between websites. And then you use these to crawl the web. Can you unpack that a little bit more?
Robert Ackland: My interest in the web has always been the fact that it’s a socially generated network of resources, the resources are web pages, and also, the other media files. The piece of engineering that connects these resources together is the hyperlink. Hyperlinks do not get formed randomly.
Hyperlinks are connecting pages together, and allowing people to, as they surf the web, find new information. This, for me has always been the thing that I’ve been most interested in, because there is a social science of why hyperlinks are created. And what does it mean for a website to create a hyperlink to another website? It’s used in order to guide people’s attention, shape people’s attention. And so the types of actors that I studied, have been political parties, social movements, organizations, activists, and they are all making choices about how to hyperlink to, and why. And these choices have measurable impacts on shaping the attention of other people.
When I started studying the web, there was not the availability of tools and techniques to allow a broad range of social scientists, particularly those with an interest in social network analysis, to easily access and collect hyperlink data, and turn these data into what I call research ready data-sets, data sets that are amenable to social network analysis, and so I really designed the VOSON software to be a tool for social network analysis using data from using hyperlink data.
The VOSON software was effectively a web crawler that allowed researchers to easily select a set of websites, and then find how those websites connected to one another through hyperlinks.
Nosh, you made you made the point that today, it’s very common to think of people networking, on the web, or via social media, But in the early days, before social media, web 1.0 was an era where you had to have quite a lot of resources, in order to be able to put material on the web, for example, newspapers or academic institutions. The typical user of the web was a consumer of information. Web 2.0, which started with blogs, but then moved on to the social media era, became an era where it was possible to not only consume information for produce information.
And so today, it’s very easy to conceptualize this idea that people go onto social media and connect with one another. In web 1.0. era, it was less easy to conceptualize this. But I really saw the hyperlink is the tool that allows organizations and groups to connect to one another. And I was interested in using social network analysis to study that phenomenon.
Noshir Contractor: So one of the things that you were pointing out is that websites are very strategic about which other websites they point to, because that’s how they represent themselves to the public, and are also very interested in which websites are pointing to them, and to the extent that we know in society that you’re judged by the friends, you keep, what you’re saying, Rob, is that a website is judged by the hyperlinks it keeps.
Robert Ackland: It’s always of great interest to know, well, who is hyperlinking, to whom? It’s a measure of popularity. In an information context, it’s a measure of authority, is your website an authoritative source on a particular topic.
It’s very important to know who is linking to you, and also, the perception of your organization is very much influenced by who you direct your hyperlinks to, and so on. It’s one of the aspects of web science that I find very interesting and compelling.
Noshir Contractor: One of the things of course, that can happen with hyperlinks, as it does today with friend links, or follow links, is that you can create them and at some point, you could dissolve them or you can unfollow someone or unfriend someone or remove a link that you have with someone. So as you look at it from a historic point of view, is it possible to be able to go back in time and look at the archive to see when someone might have created a link from one website to another and when it might have dissolved. And what that might tell us about society?
Robert Ackland: It’s a really important aspect of research in the sense that the web is constantly changing. This is one of the reasons why governments are very concerned about preserving the web, because it’s a digital record of a country or of a society.
From the perspective of a web scientist, I think there’s really two aspects of hyperlinks that are, in some ways, the holy grail for research. Number one: I find that when I present my hyperlink research to people, one of the first things they say is, you know, how has it changed over time? Another aspect that is very important, is knowing what amount of attention is traveling through hyperlink, it’s difficult to know exactly how many people were following that hyperlink.
Noshir Contractor: So one of the interesting and important contributions, Rob, that you have made to the study of web science, is the development of this virtual observatory for the study of online networks. When you began that effort, it was focused largely on mapping hyperlinks between websites, and since then you have evolved the entire project to also look for mapping links that happen between organizations or people or organizations that have Twitter accounts . Tell us a little bit about why you got interested in creating what I think has become a remarkable public good for anyone interested in studying web science.
Robert Ackland: I was always interested in developing tools that could be used by non-programmers. Web science brings people from a whole lot ot disciplines. And the whole point of web science in my mind is studying how the web is contributing to society from a lot of different dimensions. It’s not just about the engineering, but it’s about the social, political and economic impacts of the web. As the web evolved to the social media era, I wanted to make sure the VOSON software evolved.
We started then, collecting data relating to Facebook. So an early version of the VOSON software enabled research of Facebook, and of course, API’s and privacy changes on behalf of the social media companies in terms of access to data means that a tool like VOSON has to constantly be evolving as well, so VOSON’s designed to allow researchers to collect data from major social media platforms using application programming interfaces. So a lot of current tool does enable collection of Twitter network data. This is, I believe, really important for the study of political deliberation, how that is occurring on social media. And so to the extent that the social media companies continue to provide open access through to their data through API’s, then I’m very keen for the VOSON software to be a part of the web science toolkit.
Noshir Contractor: Along with your evolution of work, from hyperlinks to looking at other social media platforms, you’ve also evolved in your conceptualization of bots, where initially you could think of a bot as being a web crawler. We now know a lot about spam bots and chat bots and bots that can conduct automated high frequency trading and global financial markets. you also talk a lot about what you refer to as social bots. First of all, how do you define a social bot? And what differentiates a social bot from some of these other bots that we’ve just talked about?
Robert Ackland: So, my interest in social bots came about around the 2016 US presidential election I think the 2016 US presidential election and also the Brexit referendum in that year, really raised awareness about the potential for social media to be a vector for influence. And the influence might be coming from foreign influence operations. So, troll accounts, for example, set up by foreign governments in order to try and influence political conversations, but another area of concern related to so called social bots. The idea of intelligent agents or bots, is not new, but the 2016 US presidential election, in around that time, there was concerns about how bots were being used, in order to shape conversations on social media. I became very interested in how to understand how bots might be having an impact on political conversations on Twitter.
And so this really gets back to a very sort of a long standing and interesting question in social science research, and which is how we measure influence. Is it the case that they are influential? Because they’re very active? And they’re tweeting a lot? Or is it the case that they’re influential because the tweets somehow help to propagate particular information or raise prominence on particular themes or frames.
I think the presence and impact of bots is a core issue and potential concern for web science.
Noshir Contractor: While there is a lot of research that highlights the dangers of the risks associated with social bots, can you talk a little bit about why and how you believe that social bots can actually help promote deliberative democracy in social media?
Robert Ackland: If we think about bots, in other areas of society and the economy, they’re generally designed to be useful, in the sense that they provide information that helps people make decisions in a financial market setting, for example.
The first work that I got involved in the area of social bots was with Tim Graham. We were interested in the potential for social bots to play a positive role in political deliberation online. If you think about what political deliberation involves, it’s this idea that people are engaging with one another, often with people who do not share the same views. And they are able to develop a common set of terms and understandings about a potentially divisive social issue, and potentially changed minds, or at least come from common understanding about what the problems are.
We were interested in the idea that it might be possible for social bots to be designed to have a positive impact on political deliberation, for example, by connecting groups of people who otherwise are not connected in online conversations. One of the bots that could be designed in such a setting was what we call the bridger bot, and the idea was that such a bot might have to try to meet communities in social media, who otherwise are not connected to one another, to help promote cross community dialogue.
Another thought that we had with regards to the potential positive role of social bots was the idea that certain clusters of social media users could benefit potentially from being exposed to ideas that are different to the ones that they currently have. And so, the idea was that it was a bot that could somehow start to operate or start to be present in their conversation, participate by raising ideas that were somehow counter to what the current thinking was. However, I would like to say this, this is where I think web science is really important — because it’s one thing for social scientists to conceptualize a popper bot or bridger bot, but this is an engineering and design issue. And so this is where websites play a role in terms of connecting engineers, computer scientists and social scientists in projects that are trying to study for example, the potential positive role of social bots.
Noshir Contractor: Speaking of positive roles of bots, you and Tim, inspired by Isaac Asimov’s three laws of robotics, postulate three principles of social bots.
Robert Ackland: So the paper on social bots that I co-authored with Tim Graham was partly inspired by our common interest in studying the web from social network analysis perspective. But there are actually two literary inspirations for this work. The first inspiration is evident in the title of the paper, which is “Do Social Bots Dream of Popping the Filter Bubble.” So this was a reference to Philip K. Dix’s seminal novel, Do Androids Dream of Electric Sheep. And this is a novel that inspired the Blade Runner film. So we were interested in this idea of social bots as autonomous agents, with a purpose. And we were interested in the idea that the purpose of social bots could be a positive one, in the sense of making a positive contribution to deliberative democracy online.
However, this is an engineering problem. We’re social scientists. But we realized that the design of a social bot is an engineering task. And so another literary inspiration for this work was Isaac Asimov, who famously proposed the three principles of robotics. And so we drew on those principles.
And I want to emphasize that this is not an engineering paper that we’re proposing. In some ways it’s a thought piece. But the first principle of robotics, or our adaptation of Asimov’s principles, was that social bot must do no harm to a human being. And so how might we think of a social work creating harm to a human being? Well, by being annoying, for example, by butting into conversations where they’re not required, by creating noise in a social media conversation.
The second is that social bots must protect their own existence, except where in doing that, it that would conflict with the first one. The idea there is that a social bot has to be designed well, in the sense that it’s, it’s not annoying, it doesn’t get outed very straight away as being a bot rather than a human. Because then that can lead to people on social media platform banning it.
And then, the third principle that was adapted again, from Asimov’s three laws of robotics, was that social bots, social bots must make a significant improvement to deliberative democracy.
Noshir Contractor: That’s brilliant. I love it. Another major contribution that you’ve made to web science is the book that you published in 2013, titled, “Web, Social Science Concepts, Data and Tools for Social Scientists in the Digital Age”. Tell us a little about your thinking when you decided that you would write this book, and tell us what you’re hoping to achieve by people who would read this book.
Robert Ackland: I’ve been involved in teaching at the IU for the last 10 years now, my teaching has been in the area of social science of the internet, online research methods. Essentially, my goal in my teaching has been to equip social science students with the conceptual concepts, and the also the tools and the methodological training, to allow them to do web science, in the sense that they can work with data being generated from the web, to understand the social, political, and economic impacts of the web.
My book, really had two goals. Firstly, it was to introduce students and researchers to the web as a source of new data for studied social, political and economic behavior, the heavy emphasis on social network analysis, but also other methodological approaches.The second aspect of the book was to provide an understanding of how social scientists can contribute to the future development and pathway of the web, in order to allow the web to reach its full potential or to continue to have its full potential in terms of making a positive contribution to society.
Noshir Contractor: I’d highly recommend that book to anyone who’s interested in helping us understand how we live online, and what are the consequences of that. It has been a true delight to get a chance to catch up with you, and to hear all about the ways in which you’ve been thinking about the past of web science, the present of web science, and also the future of web science. And I’m very encouraged and inspired by everything you’ve done to contribute to the web science community in terms of your own research, in terms of the platforms like VOSON, that you have helped develop, and the book that you helped write, to help shape the next generation of students. So thank you again, Rob, for joining us today.
Robert Ackland: Thank you. It’s been my pleasure to participate in Web Science and participate in this podcast, thanks Nosh.