Brian Fung in the Washington Post: “It turns out that OkCupid has been performing some of the same psychological experiments on its users that landed Facebook in hot water recently.
In a lengthy blog post, OkCupid cofounder Christian Rudder explains that OkCupid has on occasion played around with removing text from people’s profiles, removing photos, and even telling some users they were an excellent match when in fact they were only a 30 percent match according to the company’s systems. Just to see what would happen.
OkCupid defends this behavior as something that any self-respecting Web site would do.
“OkCupid doesn’t really know what it’s doing. Neither does any other Web site,” Rudder wrote. “But guess what, everybody: if you use the Internet, you’re the subject of hundreds of experiments at any given time, on every site. That’s how websites work.”…
we have a bigger problem on our hands: A problem about how to reconcile the sometimes valuable lessons of data science with the creep factor — particularly when you aren’t notified about being studied. But as I’ve written before, these kinds of studies happen all the time; it’s just rare that the public is presented with the results.
Short of banning the practice altogether, which seems totally unrealistic, corporate data science seems like an opportunity on a number of levels, particularly if it’s disclosed to the public. First, it helps us understand how human beings tend to behave at Internet scale. Second, it tells us more about how Internet companies work. And third, it helps consumers make better decisions about which services they’re comfortable using.
I suspect that what bothers us most of all is not that the research took place, but that we’re slowly coming to grips with how easily we ceded control over our own information — and how the machines that collect all this data may all know more about us than we do ourselves. We had no idea we were even in a rabbit hole, and now we’ve discovered we’re 10 feet deep. As many as 62.5 percent of Facebook users don’t know the news feed is generated by a company algorithm, according to a recent study conducted by Christian Sandvig, an associate professor at the University of Michigan, and Karrie Karahalios, an associate professor at the University of Illinois.
OkCupid’s blog post is distinct in several ways from Facebook’s psychological experiment. OkCupid didn’t try to publish its findings in a scientific journal. It isn’t even claiming that what it did was science. Moreover, OkCupid’s research is legitimately useful to users of the service — in ways that Facebook’s research is arguably not….
But in any case, there’s no such motivating factor when it comes to Facebook. Unless you’re a page administrator or news organization, understanding how the newsfeed works doesn’t really help the average user in the way that understanding how OkCupid works does. That’s because people use Facebook for all kinds of reasons that have nothing to do with Facebook’s commercial motives. But people would stop using OkCupid if they discovered it didn’t “work.”
If you’re lying to your users in an attempt to improve your service, what’s the line between A/B testing and fraud?”
Request for Proposals: Exploring the Implications of Government Release of Large Datasets
“The Berkeley Center for Law & Technology and Microsoft are issuing this request for proposals (RFP) to fund scholarly inquiry to examine the civil rights, human rights, security and privacy issues that arise from recent initiatives to release large datasets of government information to the public for analysis and reuse. This research may help ground public policy discussions and drive the development of a framework to avoid potential abuses of this data while encouraging greater engagement and innovation.
This RFP seeks to:
- Gain knowledge of the impact of the online release of large amounts of data generated by citizens’ interactions with government
- Imagine new possibilities for technical, legal, and regulatory interventions that avoid abuse
- Begin building a body of research that addresses these issues
– BACKGROUND –
Governments at all levels are releasing large datasets for analysis by anyone for any purpose—“Open Data.” Using Open Data, entrepreneurs may create new products and services, and citizens may use it to gain insight into the government. A plethora of time saving and other useful applications have emerged from Open Data feeds, including more accurate traffic information, real-time arrival of public transportation, and information about crimes in neighborhoods. Sometimes governments release large datasets in order to encourage the development of unimagined new applications. For instance, New York City has made over 1,100 databases available, some of which contain information that can be linked to individuals, such as a parking violation database containing license plate numbers and car descriptions.
Data held by the government is often implicitly or explicitly about individuals—acting in roles that have recognized constitutional protection, such as lobbyist, signatory to a petition, or donor to a political cause; in roles that require special protection, such as victim of, witness to, or suspect in a crime; in the role as businessperson submitting proprietary information to a regulator or obtaining a business license; and in the role of ordinary citizen. While open government is often presented as an unqualified good, sometimes Open Data can identify individuals or groups, leading to a more transparent citizenry. The citizen who foresees this growing transparency may be less willing to engage in government, as these transactions may be documented and released in a dataset to anyone to use for any imaginable purpose—including to deanonymize the database—forever. Moreover, some groups of citizens may have few options or no choice as to whether to engage in governmental activities. Hence, open data sets may have a disparate impact on certain groups. The potential impact of large-scale data and analysis on civil rights is an area of growing concern. A number of civil rights and media justice groups banded together in February 2014 to endorse the “Civil Rights Principles for the Era of Big Data” and the potential of new data systems to undermine longstanding civil rights protections was flagged as a “central finding” of a recent policy review by White House adviser John Podesta.
The Berkeley Center for Law & Technology (BCLT) and Microsoft are issuing this request for proposals in an effort to better understand the implications and potential impact of the release of data related to U.S. citizens’ interactions with their local, state and federal governments. BCLT and Microsoft will fund up to six grants, with a combined total of $300,000. Grantees will be required to participate in a workshop to present and discuss their research at the Berkeley Technology Law Journal (BTLJ) Spring Symposium. All grantees’ papers will be published in a dedicated monograph. Grantees’ papers that approach the issues from a legal perspective may also be published in the BTLJ. We may also hold a followup workshop in New York City or Washington, DC.
While we are primarily interested in funding proposals that address issues related to the policy impacts of Open Data, many of these issues are intertwined with general societal implications of “big data.” As a result, proposals that explore Open Data from a big data perspective are welcome; however, proposals solely focused on big data are not. We are open to proposals that address the following difficult question. We are also open to methods and disciplines, and are particularly interested in proposals from cross-disciplinary teams.
- To what extent does existing Open Data made available by city and state governments affect individual profiling? Do the effects change depending on the level of aggregation (neighborhood vs. cities)? What releases of information could foreseeably cause discrimination in the future? Will different groups in society be disproportionately impacted by Open Data?
- Should the use of Open Data be governed by a code of conduct or subject to a review process before being released? In order to enhance citizen privacy, should governments develop guidelines to release sampled or perturbed data, instead of entire datasets? When datasets contain potentially identifiable information, should there be a notice-and-comment proceeding that includes proposed technological solutions to anonymize, de-identify or otherwise perturb the data?
- Is there something fundamentally different about government services and the government’s collection of citizen’s data for basic needs in modern society such as power and water that requires governments to exercise greater due care than commercial entities?
- Companies have legal and practical mechanisms to shield data submitted to government from public release. What mechanisms do individuals have or should have to address misuse of Open Data? Could developments in the constitutional right to information policy as articulated in Whalen and Westinghouse Electric Co address Open Data privacy issues?
- Collecting data costs money, and its release could affect civil liberties. Yet it is being given away freely, sometimes to immensely profitable firms. Should governments license data for a fee and/or impose limits on its use, given its value?
- The privacy principle of “collection limitation” is under siege, with many arguing that use restrictions will be more efficacious for protecting privacy and more workable for big data analysis. Does the potential of Open Data justify eroding state and federal privacy act collection limitation principles? What are the ethical dimensions of a government system that deprives the data subject of the ability to obscure or prevent the collection of data about a sensitive issue? A move from collection restrictions to use regulation raises a number of related issues, detailed below.
- Are use restrictions efficacious in creating accountability? Consumer reporting agencies are regulated by use restrictions, yet they are not known for their accountability. How could use regulations be implemented in the context of Open Data efficaciously? Can a self-learning algorithm honor data use restrictions?
- If an Open Dataset were regulated by a use restriction, how could individuals police wrongful uses? How would plaintiffs overcome the likely defenses or proof of facts in a use regulation system, such as a burden to prove that data were analyzed and the product of that analysis was used in a certain way to harm the plaintiff? Will plaintiffs ever be able to beat first amendment defenses?
- The President’s Council of Advisors on Science and Technology big data report emphasizes that analysis is not a “use” of data. Such an interpretation suggests that NSA metadata analysis and large-scale scanning of communications do not raise privacy issues. What are the ethical and legal implications of the “analysis is not use” argument in the context of Open Data?
- Open Data celebrates the idea that information collected by the government can be used by another person for various kinds of analysis. When analysts are not involved in the collection of data, they are less likely to understand its context and limitations. How do we ensure that this knowledge is maintained in a use regulation system?
- Former President William Clinton was admitted under a pseudonym for a procedure at a New York Hospital in 2004. The hospital detected 1,500 attempts by its own employees to access the President’s records. With snooping such a tempting activity, how could incentives be crafted to cause self-policing of government data and the self-disclosure of inappropriate uses of Open Data?
- It is clear that data privacy regulation could hamper some big data efforts. However, many examples of big data successes hail from highly regulated environments, such as health care and financial services—areas with statutory, common law, and IRB protections. What are the contours of privacy law that are compatible with big data and Open Data success and which are inherently inimical to it?
- In recent years, the problem of “too much money in politics” has been addressed with increasing disclosure requirements. Yet, distrust in government remains high, and individuals identified in donor databases have been subjected to harassment. Is the answer to problems of distrust in government even more Open Data?
- What are the ethical and epistemological implications of encouraging government decision-making based upon correlation analysis, without a rigorous understanding of cause and effect? Are there decisions that should not be left to just correlational proof? While enthusiasm for data science has increased, scientific journals are elevating their standards, with special scrutiny focused on hypothesis-free, multiple comparison analysis. What could legal and policy experts learn from experts in statistics about the nature and limits of open data?…
To submit a proposal, visit the Conference Management Toolkit (CMT) here.
Once you have created a profile, the site will allow you to submit your proposal.
If you have questions, please contact Chris Hoofnagle, principal investigator on this project.”
No silver bullet: De-identification still doesn’t work
Arvind Narayanan and Edward W. Felten: “Paul Ohm’s 2009 article Broken Promises of Privacy spurred a debate in legal and policy circles on the appropriate response to computer science research on re-identification techniques. In this debate, the empirical research has often been misunderstood or misrepresented. A new report by Ann Cavoukian and Daniel Castro is full of such inaccuracies, despite its claims of “setting the record straight.” In a response to this piece, Ed Felten and I point out eight of our most serious points of disagreement with Cavoukian and Castro. The thrust of our arguments is that (i) there is no evidence that de-identification works either in theory or in practice and (ii) attempts to quantify its efficacy are unscientific and promote a false sense of security by assuming unrealistic, artificially constrained models of what an adversary might do. Specifically, we argue that:
- There is no known effective method to anonymize location data, and no evidence that it’s meaningfully achievable.
- Computing re-identification probabilities based on proof-of-concept demonstrations is silly.
- Cavoukian and Castro ignore many realistic threats by focusing narrowly on a particular model of re-identification.
- Cavoukian and Castro concede that de-identification is inadequate for high-dimensional data. But nowadays most interesting datasets are high-dimensional.
- Penetrate-and-patch is not an option.
- Computer science knowledge is relevant and highly available.
- Cavoukian and Castro apply different standards to big data and re-identification techniques.
- Quantification of re-identification probabilities, which permeates Cavoukian and Castro’s arguments, is a fundamentally meaningless exercise.
Data privacy is a hard problem. Data custodians face a choice between roughly three alternatives: sticking with the old habit of de-identification and hoping for the best; turning to emerging technologies like differential privacy that involve some trade-offs in utility and convenience; and using legal agreements to limit the flow and use of sensitive data. These solutions aren’t fully satisfactory, either individually or in combination, nor is any one approach the best in all circumstances. Change is difficult. When faced with the challenge of fostering data science while preventing privacy risks, the urge to preserve the status quo is understandable. However, this is incompatible with the reality of re-identification science. If a “best of both worlds” solution exists, de-identification is certainly not that solution. Instead of looking for a silver bullet, policy makers must confront hard choices.”
Selected Readings on Crowdsourcing Tasks and Peer Production
The Living Library’s Selected Readings series seeks to build a knowledge base on innovative approaches for improving the effectiveness and legitimacy of governance. This curated and annotated collection of recommended works on the topic of crowdsourcing was originally published in 2014.
Technological advances are creating a new paradigm by which institutions and organizations are increasingly outsourcing tasks to an open community, allocating specific needs to a flexible, willing and dispersed workforce. “Microtasking” platforms like Amazon’s Mechanical Turk are a burgeoning source of income for individuals who contribute their time, skills and knowledge on a per-task basis. In parallel, citizen science projects – task-based initiatives in which citizens of any background can help contribute to scientific research – like Galaxy Zoo are demonstrating the ability of lay and expert citizens alike to make small, useful contributions to aid large, complex undertakings. As governing institutions seek to do more with less, looking to the success of citizen science and microtasking initiatives could provide a blueprint for engaging citizens to help accomplish difficult, time-consuming objectives at little cost. Moreover, the incredible success of peer-production projects – best exemplified by Wikipedia – instills optimism regarding the public’s willingness and ability to complete relatively small tasks that feed into a greater whole and benefit the public good. You can learn more about this new wave of “collective intelligence” by following the MIT Center for Collective Intelligence and their annual Collective Intelligence Conference.
Selected Reading List (in alphabetical order)
- Yochai Benkler — The Wealth of Networks: How Social Production Transforms Markets and Freedom — a book on the ways commons-based peer-production is transforming modern society.
- Daren C. Brabham — Using Crowdsourcing in Government — a report describing the diversity of methods crowdsourcing could be greater utilized by governments, including through the leveraging of micro-tasking platforms.
- Kevin J. Boudreau, Patrick Gaule, Karim Lakhani, Christoph Reidl, Anita Williams Woolley – From Crowds to Collaborators: Initiating Effort & Catalyzing Interactions Among Online Creative Workers – a working paper exploring the conditions,
- including incentives, that affect online collaboration.
- Chiara Franzoni and Henry Sauermann — Crowd Science: The Organization of Scientific Research in Open Collaborative Projects — a paper describing the potential advantages of deploying crowd science in a variety of contexts.
- Aniket Kittur, Ed H. Chi and Bongwon Suh — Crowdsourcing User Studies with Mechanical Turk — a paper proposing potential benefits beyond simple task completion for microtasking platforms like Mechanical Turk.
- Aniket Kittur, Jeffrey V. Nickerson, Michael S. Bernstein, Elizabeth M. Gerber, Aaron Shaw, John Zimmerman, Matthew Lease, and John J. Horton — The Future of Crowd Work — a paper describing the promise of increased and evolved crowd work’s effects on the global economy.
- Michael J. Madison — Commons at the Intersection of Peer Production, Citizen Science, and Big Data: Galaxy Zoo — an in-depth case study of the Galaxy Zoo containing insights regarding the importance of clear objectives and institutional and/or professional collaboration in citizen science initiatives.
- Thomas W. Malone, Robert Laubacher and Chrysanthos Dellarocas – Harnessing Crowds: Mapping the Genome of Collective Intelligence – an article proposing a framework for understanding collective intelligence efforts.
- Geoff Mulgan – True Collective Intelligence? A Sketch of a Possible New Field – a paper proposing theoretical building blocks and an experimental and research agenda around the field of collective intelligence.
- Henry Sauermann and Chiara Franzoni – Participation Dynamics in Crowd-Based Knowledge Production: The Scope and Sustainability of Interest-Based Motivation – a paper exploring the role of interest-based motivation in collaborative knowledge production.
- Catherine E. Schmitt-Sands and Richard J. Smith – Prospects for Online Crowdsourcing of Social Science Research Tasks: A Case Study Using Amazon Mechanical Turk – an article describing an experiment using Mechanical Turk to crowdsource public policy research microtasks.
- Clay Shirky — Here Comes Everybody: The Power of Organizing Without Organizations — a book exploring the ways largely unstructured collaboration is remaking practically all sectors of modern life.
- Jonathan Silvertown — A New Dawn for Citizen Science — a paper examining the diverse factors influencing the emerging paradigm of “science by the people.”
- Katarzyna Szkuta, Roberto Pizzicannella, David Osimo – Collaborative approaches to public sector innovation: A scoping study – an article studying success factors and incentives around the collaborative delivery of online public services.
Annotated Selected Reading List (in alphabetical order)
Benkler, Yochai. The Wealth of Networks: How Social Production Transforms Markets and Freedom. Yale University Press, 2006. http://bit.ly/1aaU7Yb.
- In this book, Benkler “describes how patterns of information, knowledge, and cultural production are changing – and shows that the way information and knowledge are made available can either limit or enlarge the ways people can create and express themselves.”
- In his discussion on Wikipedia – one of many paradigmatic examples of people collaborating without financial reward – he calls attention to the notable ongoing cooperation taking place among a diversity of individuals. He argues that, “The important point is that Wikipedia requires not only mechanical cooperation among people, but a commitment to a particular style of writing and describing concepts that is far from intuitive or natural to people. It requires self-discipline. It enforces the behavior it requires primarily through appeal to the common enterprise that the participants are engaged in…”
Brabham, Daren C. Using Crowdsourcing in Government. Collaborating Across Boundaries Series. IBM Center for The Business of Government, 2013. http://bit.ly/17gzBTA.
- In this report, Brabham categorizes government crowdsourcing cases into a “four-part, problem-based typology, encouraging government leaders and public administrators to consider these open problem-solving techniques as a way to engage the public and tackle difficult policy and administrative tasks more effectively and efficiently using online communities.”
- The proposed four-part typology describes the following types of crowdsourcing in government:
- Knowledge Discovery and Management
- Distributed Human Intelligence Tasking
- Broadcast Search
- Peer-Vetted Creative Production
- In his discussion on Distributed Human Intelligence Tasking, Brabham argues that Amazon’s Mechanical Turk and other microtasking platforms could be useful in a number of governance scenarios, including:
- Governments and scholars transcribing historical document scans
- Public health departments translating health campaign materials into foreign languages to benefit constituents who do not speak the native language
- Governments translating tax documents, school enrollment and immunization brochures, and other important materials into minority languages
- Helping governments predict citizens’ behavior, “such as for predicting their use of public transit or other services or for predicting behaviors that could inform public health practitioners and environmental policy makers”
Boudreau, Kevin J., Patrick Gaule, Karim Lakhani, Christoph Reidl, Anita Williams Woolley. “From Crowds to Collaborators: Initiating Effort & Catalyzing Interactions Among Online Creative Workers.” Harvard Business School Technology & Operations Mgt. Unit Working Paper No. 14-060. January 23, 2014. https://bit.ly/2QVmGUu.
- In this working paper, the authors explore the “conditions necessary for eliciting effort from those affecting the quality of interdependent teamwork” and “consider the the role of incentives versus social processes in catalyzing collaboration.”
- The paper’s findings are based on an experiment involving 260 individuals randomly assigned to 52 teams working toward solutions to a complex problem.
- The authors determined the level of effort in such collaborative undertakings are sensitive to cash incentives. However, collaboration among teams was driven more by the active participation of teammates, rather than any monetary reward.
Franzoni, Chiara, and Henry Sauermann. “Crowd Science: The Organization of Scientific Research in Open Collaborative Projects.” Research Policy (August 14, 2013). http://bit.ly/HihFyj.
- In this paper, the authors explore the concept of crowd science, which they define based on two important features: “participation in a project is open to a wide base of potential contributors, and intermediate inputs such as data or problem solving algorithms are made openly available.” The rationale for their study and conceptual framework is the “growing attention from the scientific community, but also policy makers, funding agencies and managers who seek to evaluate its potential benefits and challenges. Based on the experiences of early crowd science projects, the opportunities are considerable.”
- Based on the study of a number of crowd science projects – including governance-related initiatives like Patients Like Me – the authors identify a number of potential benefits in the following categories:
- Knowledge-related benefits
- Benefits from open participation
- Benefits from the open disclosure of intermediate inputs
- Motivational benefits
- The authors also identify a number of challenges:
- Organizational challenges
- Matching projects and people
- Division of labor and integration of contributions
- Project leadership
- Motivational challenges
- Sustaining contributor involvement
- Supporting a broader set of motivations
- Reconciling conflicting motivations
Kittur, Aniket, Ed H. Chi, and Bongwon Suh. “Crowdsourcing User Studies with Mechanical Turk.” In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, 453–456. CHI ’08. New York, NY, USA: ACM, 2008. http://bit.ly/1a3Op48.
- In this paper, the authors examine “[m]icro-task markets, such as Amazon’s Mechanical Turk, [which] offer a potential paradigm for engaging a large number of users for low time and monetary costs. [They] investigate the utility of a micro-task market for collecting user measurements, and discuss design considerations for developing remote micro user evaluation tasks.”
- The authors conclude that in addition to providing a means for crowdsourcing small, clearly defined, often non-skill-intensive tasks, “Micro-task markets such as Amazon’s Mechanical Turk are promising platforms for conducting a variety of user study tasks, ranging from surveys to rapid prototyping to quantitative measures. Hundreds of users can be recruited for highly interactive tasks for marginal costs within a timeframe of days or even minutes. However, special care must be taken in the design of the task, especially for user measurements that are subjective or qualitative.”
Kittur, Aniket, Jeffrey V. Nickerson, Michael S. Bernstein, Elizabeth M. Gerber, Aaron Shaw, John Zimmerman, Matthew Lease, and John J. Horton. “The Future of Crowd Work.” In 16th ACM Conference on Computer Supported Cooperative Work (CSCW 2013), 2012. http://bit.ly/1c1GJD3.
- In this paper, the authors discuss paid crowd work, which “offers remarkable opportunities for improving productivity, social mobility, and the global economy by engaging a geographically distributed workforce to complete complex tasks on demand and at scale.” However, they caution that, “it is also possible that crowd work will fail to achieve its potential, focusing on assembly-line piecework.”
- The authors argue that seven key challenges must be met to ensure that crowd work processes evolve and reach their full potential:
- Designing workflows
- Assigning tasks
- Supporting hierarchical structure
- Enabling real-time crowd work
- Supporting synchronous collaboration
- Controlling quality
Madison, Michael J. “Commons at the Intersection of Peer Production, Citizen Science, and Big Data: Galaxy Zoo.” In Convening Cultural Commons, 2013. http://bit.ly/1ih9Xzm.
- This paper explores a “case of commons governance grounded in research in modern astronomy. The case, Galaxy Zoo, is a leading example of at least three different contemporary phenomena. In the first place, Galaxy Zoo is a global citizen science project, in which volunteer non-scientists have been recruited to participate in large-scale data analysis on the Internet. In the second place, Galaxy Zoo is a highly successful example of peer production, some times known as crowdsourcing…In the third place, is a highly visible example of data-intensive science, sometimes referred to as e-science or Big Data science, by which scientific researchers develop methods to grapple with the massive volumes of digital data now available to them via modern sensing and imaging technologies.”
- Madison concludes that the success of Galaxy Zoo has not been the result of the “character of its information resources (scientific data) and rules regarding their usage,” but rather, the fact that the “community was guided from the outset by a vision of a specific organizational solution to a specific research problem in astronomy, initiated and governed, over time, by professional astronomers in collaboration with their expanding universe of volunteers.”
Malone, Thomas W., Robert Laubacher and Chrysanthos Dellarocas. “Harnessing Crowds: Mapping the Genome of Collective Intelligence.” MIT Sloan Research Paper. February 3, 2009. https://bit.ly/2SPjxTP.
- In this article, the authors describe and map the phenomenon of collective intelligence – also referred to as “radical decentralization, crowd-sourcing, wisdom of crowds, peer production, and wikinomics – which they broadly define as “groups of individuals doing things collectively that seem intelligent.”
- The article is derived from the authors’ work at MIT’s Center for Collective Intelligence, where they gathered nearly 250 examples of Web-enabled collective intelligence. To map the building blocks or “genes” of collective intelligence, the authors used two pairs of related questions:
- Who is performing the task? Why are they doing it?
- What is being accomplished? How is it being done?
- The authors concede that much work remains to be done “to identify all the different genes for collective intelligence, the conditions under which these genes are useful, and the constraints governing how they can be combined,” but they believe that their framework provides a useful start and gives managers and other institutional decisionmakers looking to take advantage of collective intelligence activities the ability to “systematically consider many possible combinations of answers to questions about Who, Why, What, and How.”
Mulgan, Geoff. “True Collective Intelligence? A Sketch of a Possible New Field.” Philosophy & Technology 27, no. 1. March 2014. http://bit.ly/1p3YSdd.
- In this paper, Mulgan explores the concept of a collective intelligence, a “much talked about but…very underdeveloped” field.
- With a particular focus on health knowledge, Mulgan “sets out some of the potential theoretical building blocks, suggests an experimental and research agenda, shows how it could be analysed within an organisation or business sector and points to possible intellectual barriers to progress.”
- He concludes that the “central message that comes from observing real intelligence is that intelligence has to be for something,” and that “turning this simple insight – the stuff of so many science fiction stories – into new theories, new technologies and new applications looks set to be one of the most exciting prospects of the next few years and may help give shape to a new discipline that helps us to be collectively intelligent about our own collective intelligence.”
Sauermann, Henry and Chiara Franzoni. “Participation Dynamics in Crowd-Based Knowledge Production: The Scope and Sustainability of Interest-Based Motivation.” SSRN Working Papers Series. November 28, 2013. http://bit.ly/1o6YB7f.
- In this paper, Sauremann and Franzoni explore the issue of interest-based motivation in crowd-based knowledge production – in particular the use of the crowd science platform Zooniverse – by drawing on “research in psychology to discuss important static and dynamic features of interest and deriv[ing] a number of research questions.”
- The authors find that interest-based motivation is often tied to a “particular object (e.g., task, project, topic)” not based on a “general trait of the person or a general characteristic of the object.” As such, they find that “most members of the installed base of users on the platform do not sign up for multiple projects, and most of those who try out a project do not return.”
- They conclude that “interest can be a powerful motivator of individuals’ contributions to crowd-based knowledge production…However, both the scope and sustainability of this interest appear to be rather limited for the large majority of contributors…At the same time, some individuals show a strong and more enduring interest to participate both within and across projects, and these contributors are ultimately responsible for much of what crowd science projects are able to accomplish.”
Schmitt-Sands, Catherine E. and Richard J. Smith. “Prospects for Online Crowdsourcing of Social Science Research Tasks: A Case Study Using Amazon Mechanical Turk.” SSRN Working Papers Series. January 9, 2014. http://bit.ly/1ugaYja.
- In this paper, the authors describe an experiment involving the nascent use of Amazon’s Mechanical Turk as a social science research tool. “While researchers have used crowdsourcing to find research subjects or classify texts, [they] used Mechanical Turk to conduct a policy scan of local government websites.”
- Schmitt-Sands and Smith found that “crowdsourcing worked well for conducting an online policy program and scan.” The microtasked workers were helpful in screening out local governments that either did not have websites or did not have the types of policies and services for which the researchers were looking. However, “if the task is complicated such that it requires ongoing supervision, then crowdsourcing is not the best solution.”
Shirky, Clay. Here Comes Everybody: The Power of Organizing Without Organizations. New York: Penguin Press, 2008. https://bit.ly/2QysNif.
- In this book, Shirky explores our current era in which, “For the first time in history, the tools for cooperating on a global scale are not solely in the hands of governments or institutions. The spread of the Internet and mobile phones are changing how people come together and get things done.”
- Discussing Wikipedia’s “spontaneous division of labor,” Shirky argues that the process is like, “the process is more like creating a coral reef, the sum of millions of individual actions, than creating a car. And the key to creating those individual actions is to hand as much freedom as possible to the average user.”
Silvertown, Jonathan. “A New Dawn for Citizen Science.” Trends in Ecology & Evolution 24, no. 9 (September 2009): 467–471. http://bit.ly/1iha6CR.
- This article discusses the move from “Science for the people,” a slogan adopted by activists in the 1970s to “’Science by the people,’ which is “a more inclusive aim, and is becoming a distinctly 21st century phenomenon.”
- Silvertown identifies three factors that are responsible for the explosion of activity in citizen science, each of which could be similarly related to the crowdsourcing of skills by governing institutions:
- “First is the existence of easily available technical tools for disseminating information about products and gathering data from the public.
- A second factor driving the growth of citizen science is the increasing realisation among professional scientists that the public represent a free source of labour, skills, computational power and even finance.
- Third, citizen science is likely to benefit from the condition that research funders such as the National Science Foundation in the USA and the Natural Environment Research Council in the UK now impose upon every grantholder to undertake project-related science outreach. This is outreach as a form of public accountability.”
Szkuta, Katarzyna, Roberto Pizzicannella, David Osimo. “Collaborative approaches to public sector innovation: A scoping study.” Telecommunications Policy. 2014. http://bit.ly/1oBg9GY.
- In this article, the authors explore cases where government collaboratively delivers online public services, with a focus on success factors and “incentives for services providers, citizens as users and public administration.”
- The authors focus on six types of collaborative governance projects:
- Services initiated by government built on government data;
- Services initiated by government and making use of citizens’ data;
- Services initiated by civil society built on open government data;
- Collaborative e-government services; and
- Services run by civil society and based on citizen data.
- The cases explored “are all designed in the way that effectively harnesses the citizens’ potential. Services susceptible to collaboration are those that require computing efforts, i.e. many non-complicated tasks (e.g. citizen science projects – Zooniverse) or citizens’ free time in general (e.g. time banks). Those services also profit from unique citizens’ skills and their propensity to share their competencies.”
The Field Guide to Data Science
Booz Allen Hamilton: “Data Science is the competitive advantage of the future for organizations interested in turning their data into a product through analytics. Industries from health, to national security, to finance, to energy can be improved by creating better data analytics through Data Science. The winners and the losers in the emerging data economy are going to be determined by their Data Science teams.
Booz Allen Hamilton created The Field Guide to Data Science to help organizations of all types and missions understand how to make use of data as a resource. The text spells out what Data Science is and why it matters to organizations as well as how to create Data Science teams. Along the way, our team of experts provides field-tested approaches, personal tips and tricks, and real-life case studies. Senior leaders will walk away with a deeper understanding of the concepts at the heart of Data Science. Practitioners will add to their toolboxes.
In The Field Guide to Data Science, our Booz Allen experts provide their insights in the following areas:
- Start Here for the Basics provides an introduction to Data Science, including what makes Data Science unique from other analysis approaches. We will help you understand Data Science maturity within an organization and how to create a robust Data Science capability.
- Take Off the Training Wheels is the practitioners guide to Data Science. We share our established processes, including our approach to decomposing complex Data Science problems, the Fractal Analytic Model. We conclude with the Guide to Analytic Selection to help you select the right analytic techniques to conquer your toughest challenges.
- Life in the Trenches gives a first hand account of life as a Data Scientist. We share insights on a variety of Data Science topics through illustrative case studies. We provide tips and tricks from our own experiences on these real-life analytic challenges.
- Putting it All Together highlights our successes creating Data Science solutions for our clients. It follows several projects from data to insights and see the impact Data Science can have on your organization…”
The Power to Decide
Special Report by Antonio Regalado in MIT Technology Review: “Back in 1956, an engineer and a mathematician, William Fair and Earl Isaac, pooled $800 to start a company. Their idea: a score to handicap whether a borrower would repay a loan.
It was all done with pen and paper. Income, gender, and occupation produced numbers that amounted to a prediction about a person’s behavior. By the 1980s the three-digit scores were calculated on computers and instead took account of a person’s actual credit history. Today, Fair Isaac Corp., or FICO, generates about 10 billion credit scores annually, calculating 50 times a year for many Americans.
This machinery hums in the background of our financial lives, so it’s easy to forget that the choice of whether to lend used to be made by a bank manager who knew a man by his handshake. Fair and Isaac understood that all this could change, and that their company didn’t merely sell numbers. “We sell a radically different way of making decisions that flies in the face of tradition,” Fair once said.
This anecdote suggests a way of understanding the era of “big data”—terabytes of information from sensors or social networks, new computer architectures, and clever software. But even supercharged data needs a job to do, and that job is always about a decision.
In this business report, MIT Technology Review explores a big question: how are data and the analytical tools to manipulate it changing decision making today? On Nasdaq, trading bots exchange a billion shares a day. Online, advertisers bid on hundreds of thousands of keywords a minute, in deals greased by heuristic solutions and optimization models rather than two-martini lunches. The number of variables and the speed and volume of transactions are just too much for human decision makers.
When there’s a person in the loop, technology takes a softer approach (see “Software That Augments Human Thinking”). Think of recommendation engines on the Web that suggest products to buy or friends to catch up with. This works because Internet companies maintain statistical models of each of us, our likes and habits, and use them to decide what we see. In this report, we check in with LinkedIn, which maintains the world’s largest database of résumés—more than 200 million of them. One of its newest offerings is University Pages, which crunches résumé data to offer students predictions about where they’ll end up working depending on what college they go to (see “LinkedIn Offers College Choices by the Numbers”).
These smart systems, and their impact, are prosaic next to what’s planned. Take IBM. The company is pouring $1 billion into its Watson computer system, the one that answered questions correctly on the game show Jeopardy! IBM now imagines computers that can carry on intelligent phone calls with customers, or provide expert recommendations after digesting doctors’ notes. IBM wants to provide “cognitive services”—computers that think, or seem to (see “Facing Doubters, IBM Expands Plans for Watson”).
Andrew Jennings, chief analytics officer for FICO, says automating human decisions is only half the story. Credit scores had another major impact. They gave lenders a new way to measure the state of their portfolios—and to adjust them by balancing riskier loan recipients with safer ones. Now, as other industries get exposed to predictive data, their approach to business strategy is changing, too. In this report, we look at one technique that’s spreading on the Web, called A/B testing. It’s a simple tactic—put up two versions of a Web page and see which one performs better (see “Seeking Edge, Websites Turn to Experiments” and “Startups Embrace a Way to Fail Fast”).
Until recently, such optimization was practiced only by the largest Internet companies. Now, nearly any website can do it. Jennings calls this phenomenon “systematic experimentation” and says it will be a feature of the smartest companies. They will have teams constantly probing the world, trying to learn its shifting rules and deciding on strategies to adapt. “Winners and losers in analytic battles will not be determined simply by which organization has access to more data or which organization has more money,” Jennings has said.
Of course, there’s danger in letting the data decide too much. In this report, Duncan Watts, a Microsoft researcher specializing in social networks, outlines an approach to decision making that avoids the dangers of gut instinct as well as the pitfalls of slavishly obeying data. In short, Watts argues, businesses need to adopt the scientific method (see “Scientific Thinking in Business”).
To do that, they have been hiring a highly trained breed of business skeptics called data scientists. These are the people who create the databases, build the models, reveal the trends, and, increasingly, author the products. And their influence is growing in business. This could be why data science has been called “the sexiest job of the 21st century.” It’s not because mathematics or spreadsheets are particularly attractive. It’s because making decisions is powerful…”
How should we analyse our lives?
Gillian Tett in the Financial Times on the challenge of using the new form of data science: “A few years ago, Alex “Sandy” Pentland, a professor of computational social sciences at MIT Media Lab, conducted a curious experiment at a Bank of America call centre in Rhode Island. He fitted 80 employees with biometric devices to track all their movements, physical conversations and email interactions for six weeks, and then used a computer to analyse “some 10 gigabytes of behaviour data”, as he recalls.
The results showed that the workers were isolated from each other, partly because at this call centre, like others of its ilk, the staff took their breaks in rotation so that the phones were constantly manned. In response, Bank of America decided to change its system to enable staff to hang out together over coffee and swap ideas in an unstructured way. Almost immediately there was a dramatic improvement in performance. “The average call-handle time decreased sharply, which means that the employees were much more productive,” Pentland writes in his forthcoming book Social Physics. “[So] the call centre management staff converted the break structure of all their call centres to this new system and forecast a $15m per year productivity increase.”
When I first heard Pentland relate this tale, I was tempted to give a loud cheer on behalf of all long-suffering call centre staff and corporate drones. Pentland’s data essentially give credibility to a point that many people know instinctively: that it is horribly dispiriting – and unproductive – to have to toil in a tiny isolated cubicle by yourself all day. Bank of America deserves credit both for letting Pentland’s team engage in this people-watching – and for changing its coffee-break schedule in response.
But there is a bigger issue at stake here too: namely how academics such as Pentland analyse our lives. We have known for centuries that cultural and social dynamics influence how we behave but until now academics could usually only measure this by looking at micro-level data, which were often subjective. Anthropology (a discipline I know well) is a case in point: anthropologists typically study cultures by painstakingly observing small groups of people and then extrapolating this in a subjective manner.
Pentland and others like him are now convinced that the great academic divide between “hard” and “soft” sciences is set to disappear, since researchers these days can gather massive volumes of data about human behaviour with precision. Sometimes this information is volunteered by individuals, on sites such as Facebook; sometimes it can be gathered from the electronic traces – the “digital breadcrumbs” – that we all deposit (when we use a mobile phone, say) or deliberately collected with biometric devices like the ones used at Bank of America. Either way, it can enable academics to monitor and forecast social interaction in a manner we could never have dreamed of before. “Social physics helps us understand how ideas flow from person to person . . . and ends up shaping the norms, productivity and creative output of our companies, cities and societies,” writes Pentland. “Just as the goal of traditional physics is to understand how the flow of energy translates into change in motion, social physics seems to understand how the flow of ideas and information translates into changes in behaviour….
But perhaps the most important point is this: whether you love or hate this new form of data science, the genie cannot be put back in the bottle. The experiments that Pentland and many others are conducting at call centres, offices and other institutions across America are simply the leading edge of a trend.
The only question now is whether these powerful new tools will be mostly used for good (to predict traffic queues or flu epidemics) or for more malevolent ends (to enable companies to flog needless goods, say, or for government control). Sadly, “social physics” and data crunching don’t offer any prediction on this issue, even though it is one of the dominant questions of our age.”
From Faith-Based to Evidence-Based: The Open Data 500 and Understanding How Open Data Helps the American Economy
Beth Noveck in Forbes: “Public funds have, after all, paid for their collection, and the law says that federal government data are not protected by copyright. By the end of 2009, the US and the UK had the only two open data one-stop websites where agencies could post and citizens could find open data. Now there are over 300 such portals for government data around the world with over 1 million available datasets. This kind of Open Data — including weather, safety and public health information as well as information about government spending — can serve the country by increasing government efficiency, shedding light on regulated industries, and driving innovation and job creation.
It’s becoming clear that open data has the potential to improve people’s lives. With huge advances in data science, we can take this data and turn it into tools that help people choose a safer hospital, pick a better place to live, improve the performance of their farm or business by having better climate models, and know more about the companies with whom they are doing business. Done right, people can even contribute data back, giving everyone a better understanding, for example of nuclear contamination in post-Fukushima Japan or incidences of price gouging in America’s inner cities.
The promise of open data is limitless. (see the GovLab index for stats on open data) But it’s important to back up our faith with real evidence of what works. Last September the GovLab began the Open Data 500 project, funded by the John S. and James L. Knight Foundation, to study the economic value of government Open Data extensively and rigorously. A recent McKinsey study pegged the annual global value of Open Data (including free data from sources other than government), at $3 trillion a year or more. We’re digging in and talking to those companies that use Open Data as a key part of their business model. We want to understand whether and how open data is contributing to the creation of new jobs, the development of scientific and other innovations, and adding to the economy. We also want to know what government can do better to help industries that want high quality, reliable, up-to-date information that government can supply. Of those 1 million datasets, for example, 96% are not updated on a regular basis.
The GovLab just published an initial working list of 500 American companies that we believe to be using open government data extensively. We’ve also posted in-depth profiles of 50 of them — a sample of the kind of information that will be available when the first annual Open Data 500 study is published in early 2014. We are also starting a similar study for the UK and Europe.
Even at this early stage, we are learning that Open Data is a valuable resource. As my colleague Joel Gurin, author of Open Data Now: the Secret to Hot Start-Ups, Smart Investing, Savvy Marketing and Fast Innovation, who directs the project, put it, “Open Data is a versatile and powerful economic driver in the U.S. for new and existing businesses around the country, in a variety of ways, and across many sectors. The diversity of these companies in the kinds of data they use, the way they use it, their locations, and their business models is one of the most striking things about our findings so far.” Companies are paradoxically building value-added businesses on top of public data that anyone can access for free….”
FULL article can be found here.
The Postmodernity of Big Data
Essay by Michael Pepi in the New Inquiry: “Big Data fascinates because its presence has always been with us in nature. Each tree, drop of rain, and the path of each grain of sand, both responds to and creates millions of data points, even on a short journey. Nature is the original algorithm, the most efficient and powerful. Mathematicians since the ancients have looked to it for inspiration; techno-capitalists now look to unlock its mysteries for private gain. Playing God has become all the more brisk and profitable thanks to cloud computing.
But beyond economic motivations for Big Data’s rise, are there also epistemological ones? Has Big Data come to try to fill the vacuum of certainty left by postmodernism? Does data science address the insecurities of the postmodern thought?
It turns out that trying to explain Big Data is like trying to explain postmodernism. Neither can be summarized effectively in a phrase, despite their champions’ efforts. Broad epistemological developments are compressed into cursory, ex post facto descriptions. Attempts to define Big Data, such as IBM’s marketing copy, which promises “insights gleaned” from “enterprise data warehouses that implement massively parallel processing,” “real-time scalability” and “parsing structured and unstructured sources,” focus on its implementation at the expense of its substance, decontextualizing it entirely . Similarly, definitions of postmodernism, like art critic Thomas McEvilley’s claim that it is “a renunciation that involves recognition of the relativity of the self—of one’s habit systems, their tininess, silliness, and arbitrariness” are accurate but abstract to the point of vagueness….
Big Data might come to be understood as Big Postmodernism: the period in which the influx of unstructured, non-teleological, non-narrative inputs ceased to destabilize the existing order but was instead finally mastered processed by sufficiently complex, distributed, and pluralized algorithmic regime. If Big Data has a skepticism built in, how this is different from the skepticism of postmodernism is perhaps impossible to yet comprehend”.
The Flow of Technology Talent into Government and Civil Society
Report for the Ford Foundation and the MacArthur Foundation: “As information technology further suffuses every aspect of our lives, government will inevitably have a role to play in ensuring that technology serves the public interest. The ability for government to improve operations and provide services to citizens more efficiently through the effective use of technology is among the greatest contemporary opportunities for the public sector…
Among the key findings of this report:
- The Current Pipeline Is Insufficient: the vast majority of interviewees indicated that there is a severe paucity of individuals with technical skills in computer science, data science, and the Internet or other information technology expertise in civil society and government. In particular, many of those interviewed noted that existing talent levels fail to meet current needs to develop, leverage, or understand technology.
- Barriers to Recruitment and Retention Are Acute: many of those interviewed said that substantial barriers thwart the effective recruitment and retention of individuals with the requisite skills in government and civil society. Among the most common barriers mentioned were those of compensation, an inability to pursue groundbreaking work, and a culture that is averse to hiring and utilizing potentially disruptive innovators
- A Major Gap between the Public-Interest and For-Profit Sectors Persists: as a related matter, interviewees discussed superior for-profit recruitment and retention models. Specifically, the for-profit sector was perceived as providing both more attractive compensation (especially to young talent) and fostering a culture of innovation, openness, and creativity that was seen as more appealing to technologists and innovators.
- A Need to Examine Models from Other Fields: interviewees noted significant space to develop new models to improve the robustness of the talent pipeline; in part, many existing models were regarded as unsustainable or incomplete. Interviewees did, however, highlight approaches from other fields that could provide relevant lessons to help guide investments in improving this pipeline.
- Significant Opportunity for Connection and Training: despite consonance among those interviewed that the pipeline was incomplete, many individuals indicated the possibility for improved and more systematic efforts to expose young technologists to public interest issues and connect them to government and civil society careers through internships, fellowships, and other training and recruitment tools.
- Culture Change Necessary: the culture of government and civil society –and its effects on recruitment and other bureaucratic processes –was seen as a vital challenge that would need to be addressed to improve the pipeline. This view manifested through comments that government and civil society organizations needed to become more open to utilizing technology and adopting a mindset of experimentation….”