Task Group Summary 5
How can social networks aid our understanding of complexity?

CHALLENGE SUMMARY

Society represents one of the most complex systems that science has ever encountered. Its six billion individuals interact with varying frequencies, generating a complex social network that plays a key role in the spread of ideas and political systems, the breakout of riots and wars, and the health and well being of most individuals. Despite the obvious scientific and economic importance of understanding social networks, today we know more about E. coli bacteria than about patterns of human interaction, partly because bacteria do not get annoyed when we put them under the microscope. Most of the research on social networks focuses on snapshots of small scale networks; there is a general paucity of research on the dynamics of large scale human interaction. Our lack of understanding of macro social networks and human behavior is not due to the lack of technologies to collect the relevant data: thanks to the computerization of most aspects of life, today an increasing amount of information is automatically collected about all of us. Mobile phone companies know who calls whom and where their consumers are; email providers keep detailed records of their consumers’ electronic communications; credit card companies and banks can piece together not only their consumers’ wealth and spending patterns, but also their travel patterns. Taken together, society is the only complex system whose components are constantly monitored, offering a potential testing ground for all complex systems theories of quantitative predictive power. However, a major problem is that most of the collected data are owned by private organizations and protected by layers of confidentiality agreements



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 41
Task Group Summary 5 How can social networks aid our understanding of complexity? CHALLENGE SUMMARY Society represents one of the most complex systems that science has ever encountered. Its six billion individuals interact with varying frequen- cies, generating a complex social network that plays a key role in the spread of ideas and political systems, the breakout of riots and wars, and the health and well being of most individuals. Despite the obvious scientific and economic importance of understanding social networks, today we know more about E. coli bacteria than about patterns of human interaction, partly because bacteria do not get annoyed when we put them under the microscope. Most of the research on social networks focuses on snapshots of small scale networks; there is a general paucity of research on the dynam- ics of large scale human interaction. Our lack of understanding of macro social networks and human behavior is not due to the lack of technologies to collect the relevant data: thanks to the computerization of most aspects of life, today an increasing amount of information is automatically collected about all of us. Mobile phone companies know who calls whom and where their consumers are; email providers keep detailed records of their consum- ers’ electronic communications; credit card companies and banks can piece together not only their consumers’ wealth and spending patterns, but also their travel patterns. Taken together, society is the only complex system whose components are constantly monitored, offering a potential testing ground for all complex systems theories of quantitative predictive power. However, a major problem is that most of the collected data are owned by private organizations and protected by layers of confidentiality agreements 

OCR for page 41
 comPlEX SYSTEmS and economic interests that keep scientists at arm’s reach from making use of them. (Ideas about privacy and the technologies that protect it are a topic in themselves.) While for decades complexity was driven by theoretical ideas, the current availability of large quantities of data is rapidly turning complexity into an empirical science, offering significant opportunities for complexity to show its relevance to society. The lack of access to relevant datasets demonstrates that the scientific community has failed to explain the scientific and societal benefits of un- derstanding human behavior. The average person understands the need to invest in biomedical research, expecting that the results will lead to better approaches to the prevention or treatment of disease. Most people also comprehend the need to study materials science, which may result in better computers and phones. Yet society in general does not recognize the value of research on data that are already collected on human behavior because there is no ready understanding of its potential value but clear concern about in- vasion of our privacy. Credit card data are an example. If researchers could collect anonymous data, that is without names or credit card numbers but with data about geographic location, age, and ethnicity it would be possible to develop significant understanding of buying habits and purchasing pat- terns. For scientists to have access to these datasets, they will need to present a compelling explanation of the scientific benefits of these datasets. Much of the discourse in the scientific community has focused on the problems surrounding user confidentiality and the adverse effects of such data col- lection processes, rather than the benefits of a systematic program in social complexity. Consequently, investment in quantitative social sciences repre- sents a tiny fraction of the current federal research budget and it is virtually not existent in the industry. Key Questions The challenge to this working group is identify one or a few research areas that would clearly illustrate the societal benefits of getting access to the currently collected high quality datasets on patterns of human activity, spelling out the societal benefits of this research in a fashion that will be obvious to taxpayers, decision-makers and database owners alike. • Identify one or several key questions pertaining to social networks and human behavior that are of fundamental importance for complexity science and offer significant potential societal payoffs.

OCR for page 41
 TaSk GRoUP SUmmaRY  • Discuss how such a proposal should be presented to the society and decision-makers to address and manage the privacy needs of individuals, or to convince these constituencies that the benefits outweigh the risks involved in such research. • Identify the type of data necessary to address the proposed question, and the private or government sources that either own or have access to useful datasets. • Identify the magnitude of the investment and the ideal mechanisms necessary to make the research program a reality. • Explore to what degree such a program would answer questions that benefit not only the understanding of social systems, but also uncover laws and mechanisms that other systems of comparable organization and complexity. Required Reading Butler D. Data sharing threatens privacy. nature 2007;449:644-645. Moody J. The importance of relationship timing for diffusion: indirect connectivity and STD infection risk. Social Forces 2002;81:25-56. Onnela J-P, et al. Structure and tie strengths in mobile communication networks. Proc natl acad Sci USa 2007;104(18):7332-336). Duncan J. Watts. A twenty-first century science, nature 2007;445-489. [Accessed online August 1, 2008: [http://www.nature.com/nature/journal/v445/n7127/full/445489a. html].] Suggested Reading Palla G, Barabási A-L, Vicsek T. Quantifying social group evolution. nature 2007;446:664- 667. Putnam R. Bowling alone: America’s declining social capital. [Accessed online June24,2008: http://xroads.virginia.edu/~hyper/DETOC/assoc/bowling.html.] [Published online on April 24, 2007, 10.1073/pnas.0610245104 2007.] Stark D, Vedres B. Social times of network spaces: Network sequences and foreign investment in Hungary. american Journal of Sociology 2006;111(5):1368-1411. TASK GROUP MEMBERS • Timothy B. Buchman, Washington University in St. Louis • Kee Chan, Boston University • Noshir S. Contractor, Northwestern University • Nathan Eagle, MIT/Santa Fe Institute

OCR for page 41
 comPlEX SYSTEmS • Joshua Epstein, The Brookings Institution • Robert A. Greenes, Arizona State University • Alan J. Hurd, Los Alamos National Laboratory • Cristopher Moore, University of New Mexico • Joshua Plotkin, University of Pennsylvania • Richard Puddy, Centers for Disease Control and Prevention • Jordan Sarver, University of Georgia TASK GROUP SUMMARY By Jordan Sarver, Graduate Science Writing Student, University of Georgia The human body is full of complex systems. The nervous system, the circulatory system, and even the immune system consist of a series of smaller components working together in the body. For years, scientists have been able to view many of the networks that exist inside the body. On a macro-scale, human relationships are an example of smaller components working together to produce a larger result as well. Disease epidemics, religions, and even worldviews are spread through human interaction. Un- like body systems, there is no tool capable of viewing all of the connections that a person makes over time; nor is there a method known to follow the propagation of an idea throughout a community. At the 2008 meeting of the National Academies Keck Futures Initiative Conference on Complex Systems, a multidisciplinary group of researchers focused their energy on the complex system of social networks. The personal connections that people create and destroy everyday form what are known as social networks which influence every aspect of our daily lives from health to knowledge. When someone makes a new friend their network increases. A new friend provides access not only to new informa- tion, but also other contacts as they meet people within someone else’s network. An example of how social networks can affect a person’s health is the flu. Imagine a passenger on an airplane with the flu virus—everyone on the airplane is exposed to this virus as well as anyone who comes into contact with the passenger throughout his travels. The spread of the flu virus is a direct result of the contact the carrier has with other people throughout his day. The group set out to find a way to interpret and map the networks that exist between people, places, and ideas. First, the group wanted to figure

OCR for page 41
 TaSk GRoUP SUmmaRY  out a way to take the information that is already public and available and then construct a network connecting people to places and beliefs. Some in the group believed the problem is that scientists have not sifted through current information properly and acquiring more would only complicate current advancements. Another problem encountered when studying social networks is the dynamic nature of social systems. People are constantly making and losing connections with other people and ideologies. Constant changes realign the threads that create social networks and makes mapping the network that much harder. Technology has become an ally in the efforts to map social networks. The Internet has become an optimum tool for monitoring human behavior. Websites like Facebook and MySpace track whom people befriend and to which organizations they belong. Accordingly group members discussed the possibility of being able to map connections and the relationships that exist between people in society using the web as a tool. Social websites pro- vide an easier way to look at some complex systems that are formed all the time. These websites could potentially be used as a tool to track the path of information through cyberspace. Topics and questions that tracked the flow of conversation included how information is disseminated throughout networks (Internet, email), why some people are superspreaders of information or disease, what net- work structures are required for optimization, how innovations are diffused over the Internet, and what role team science plays in social networking. Not all information is easily accessible. Americans are extremely private people. Media has allowed Americans to watch, read, and listen to things that interest them without being subject to public scrutiny. In turn, an en- vironment where privacy is extremely valued has been created. Several types of information that are currently protected by privacy laws could potentially reveal helpful information. Medical records are kept in strict confidence by healthcare professionals, but group members noted the value of linking the spread of disease with medical records. Constructing a network of disease transmission could not only lead to more effective treatment, but it could also provide information that will allow for preemptive measures to halt the spread of certain diseases. Many of the group members brought along their notebook computers and used various websites as sources of information about social networks. The alleged link between autism and childhood vaccinations was a case study used by the group to look at how a specific idea flows through cyber- space. Using Google Trends, a subset of the Google search engine, group

OCR for page 41
 comPlEX SYSTEmS members were able to see how many Internet users searched the Internet for a link between autism and vaccinations. The website provided specific infor- mation about which states searched for the topic of autism and vaccinations the most and which websites were the most viewed on the subject. Information from the Centers for Disease Control and Prevention along with the Google Trends data allowed the group members to construct a preliminary relationship between Internet searches and real-world ef- fects. Group members examined data from Iowa where there was a mumps outbreak in 2006 and the relationship, if any, to the belief that childhood vaccinations cause autism. The lack of vaccinations appeared to correspond with the outbreak, but there was no concrete evidence available to prove that the lack of vaccinations was a result of the belief that vaccinations cause autism. The ability to track the spread of opinion with the spread of disease could potentially allow those in the field of prevention to be more specific in targeting at risk populations. Another example included the recent rise and spread of “direct-to- consumer genetic testing” as a potential topic to explore further. Both text- mining and web crawling were highlighted as potential techniques to explore multi-dimensional networks using such sites as the “thewaybackmachine,” “Digg,” “Xobni,” “dotnetmap,” “Google Trends,” “Facebook,” etc. Another important aspect of social networks is the idea of team science. Members of the group were intrigued by the thought of creating a way to predict likely collaborations among scientists across disciplines. In fact, it was noted by one of the group members that articles with the highest impact include multiple authors across several disciplines. An article will have an even higher impact if those authors from different disciplines are geographically diverse. What was developed in the group was the idea of creating a way to predict collaboration using subject matter of an author’s previous work, proximity to other scientists, and previous collaborators. Naturally, group members decided that more information is needed. Medical records are highly privatized, but full of rich information, and so the issue is whether scientists can encourage people to be more open with their private information in the interest of creating large, medically useful databases for the common good. An equally important goal is to encourage the exchange of information among health professionals who have a need to know. What are the big picture ideas? Constructing a way to facilitate team science across a multi-disciplinary field was a major priority of this group. These group members from different disciplines are an example of the

OCR for page 41
 TaSk GRoUP SUmmaRY  extraordinary results that can be reached when different scientific minds collaborate. Mapping the propagation of an idea or an action with the propagation of a disease is now possible using the information available. Also, human behavior influences disease trends which in turn, induce more human behavior. The group members realized the capabilities of the Internet to take snapshots of trends in both cyberspace and the real-world. By assembling these snapshots, scientists can produce a moving picture and watch the progression of either disease or ideologies across time and space. The next step for the group was developing a way to construct a net- work and watch it grow. Facebook, the social website designed for college students, seemed like an optimum way to watch the creation and spread of a network. Various applications exist on the Facebook website. These applica- tions allow members of the website to play games or connect with friends through shared interests. The group proposed tracing the application that allows members to either become a vampire or a werewolf. Members can then bite friends, making them a member of either the vampire or werewolf group. Although the application is fairly innocuous, the information could be useful. An application like the one on Facebook could predict trends. Members of a group who are influential and the conditions that predict who will become a part of a new network could all be revealed using the simple application. The results could be wide reaching in terms of predicting how information will spread through cyberspace. It could also predict how disease will progress by determining who is likely to be a super-spreader, someone who interacts with numerous amounts of people. By identifying super-spreaders and who they will potential interact with, prevention could become more efficient. The Internet is both a source of information and a tool to decipher information. Social science has potentially reached the end of a paradigm that has existed for a long time—the lack of understanding of contact pat- terns of populations. Now there is data to address the challenges that the field has faced. Flow can now be visualized. Scientists are no longer relegated to watching flow lines and calculating equations. The capacity to examine minute by minute construction of our world through relationships is now possible. Other research priorities include: computational thinking, making the implicit explicit, capturing relational data on a massive scale, accounting for visualizations and dynamics in this new environment, moving society from a snapshot to a moving picture of social networks, and investigating how ideas spread as well as how to spread ideas.

OCR for page 41