Crowdsourcing and Gov 2.0Irene Ros, Yannick Assogba, Joan DiMicco IBM Research One Rogers St, Cambridge, 02142 iros; yannick; joan.dimiccous.ibm.comINTRODUCTION The openness and proliferation of the social web inspired a movement known as Gov 2.0 or Open Government. The Gov 2.0 movement strives to make our government as accessible, transparent and collaborative as the services being used by Internet users on a daily basis. There are many definitions of Gov 2.0 3; Some focus on real time data availability and some argue the focus should be on using the principles exhibited by the emerging tools that only an outsider can function as a “watchdog”, supervising the activity and data generated by our government representatives. Despite the challenges, President Obamas memorandum on Open Government 10 called on the US government to become more transparent and open.THE OPEN GOVERNMENT DATA AND SOCIAL WEB MOVEMENTS Many government offices answered the Presidents call, releasing copious amount of data. This data, often available as single file dumps, or worse - html only interactive tables, are difficult to turn into actual insight. What are the ramifications of this data flood? First, it offers the illusion that the government is now transparent and its operations can be easily discerned from the data. This is far from the truth: citizens are not equipped or motivated to parse large data dumps and build their conclusions upon it, Much of the data being released is of little importance to the majority of citizens and some have suggested that this data is released merely for the sake of appearing transparent 4. These issues alongside those mentioned above motivate a community of developers who wish to convert the data into meaningful interactions for citizens, as indicated by examples such as “Code for America” 5 and “Apps for America” 2. Many applications are being built on top of data, some utilizing concepts we know from the social web such as commenting, sharing, ranking, embedding and remixing. A prominent example being OpenCongress.org which offers a robust commenting system on the contents of legislative bills. Are the sorts of applications being created today sufficient for meeting the goals of the gov 2.0 movement? How can, and does crowdsourcing play a role in the gov 2.0 ecosystem?GENRES OF CROWDSOURCING AND THEIR APPLIC- ABILITY TO THE OPEN GOVERNMENT SPACE Quinn and Bederson define a taxonomy of distributed human labor genres and applications 11. These include:Games with a purpose: Applications that employ game mechanics to encourage volunteers to perform computation. The focus is to make the game fun enough that volunteers enjoy playing it while performing work as a side effect.Mechanized Labor: Crowdsourcing applications that involved monetary rewards for human labor. The most notable example being Amazons Mechanical Turk 1.Wisdom of Crowds: The collective intelligence obtained by a distributed group of people thinking independently. 12Dual-purpose Work: This type of work is characterized by coupling computation with something else that the user is required to do, an example of this is ReCaptcha 13.Grand Search: Work that utilizes volunteers to search through a large data space to retrieve a single solution.Human-based Genetic Algorithms: An approach where the solutions consist of a series of small tasks that evolve based on human evaluation.Knowledge Collection from Volunteer Contributors - This is using human workers to build large datasets on a common topic.While the space of approaches is large, some are more difficult to apply to the Gov 2.0 space successfully due to the nature of the work in question. For example, creating a 1game around government data that offers sufficient entertainment to satisfy players is quite challenging. In the case of mechanized labor, which can be applied to almost any situation where the task can be split up efficiently, when it comes to understanding the operations of the government, some tasks require prior knowledge and context to be executed accurately. Given that open government work is often not a revenue generating business, budgeting for mechanized labor is another challenge. Lastly, tasks that aggregate the wisdom of the crowd to create a knowledge set or rely on volunteers may be at risk of inaccuracy due to contamination by political bias. Citizens participating in online political activities chose to do so to promote their view points; when collecting information from these citizens, the resulting data can be skewed by position or political leanings. When obtaining input from citizens about government decisions and operation, this form of input may be very valuable, but when attempting to derive a collective knowledge that is impartial, it can hamper ones effort. A familiar example of this is Wikipedias edit wars on controversial topics.There are