Big data From Wikipedia, the free encyclopedia Jump to: navigation, search A visualization created by IBM of Wikipedia

edits. At multiple terabytes in size , the text and images of Wikipedia are a classic example of big data. Big data[1][2] is a collection of data sets so large and complex that it becomes difficult to process using on-hand database management tools or traditional dat a processing applications. The challenges include capture, curation, storage,[3] search, sharing, transfer, analysis,[4] and visualization. The trend to larger data sets is due to the additional information derivable from analysis of a sing le large set of related data, as compared to separate smaller sets with the same total amount of data, allowing correlations to be found to "spot business trend s, determine quality of research, prevent diseases, link legal citations, combat crime, and determine real-time roadway traffic conditions."[5][6][7] As of 2012, limits on the size of data sets that are feasible to process in a re asonable amount of time were on the order of exabytes of data.[8][9] Scientists regularly encounter limitations due to large data sets in many areas, including meteorology, genomics,[10] connectomics, complex physics simulations,[11] and bi ological and environmental research.[12] The limitations also affect Internet se arch, finance and business informatics. Data sets grow in size in part because t hey are increasingly being gathered by ubiquitous information-sensing mobile dev ices, aerial sensory technologies (remote sensing), software logs, cameras, micr ophones, radio-frequency identification readers, and wireless sensor networks.[1 3][14] The world's technological per-capita capacity to store information has ro ughly doubled every 40 months since the 1980s;[15] as of 2012, every day 2.5 qui ntillion (2.51018) bytes of data were created.[16] The challenge for large enterp rises is determining who should own big data initiatives that straddle the entir e organization.[17] Big data is difficult to work with using most relational database management sys tems and desktop statistics and visualization packages, requiring instead "massi vely parallel software running on tens, hundreds, or even thousands of servers". [18] What is considered "big data" varies depending on the capabilities of the o rganization managing the set, and on the capabilities of the applications that a re traditionally used to process and analyze the data set in its domain. "For so me organizations, facing hundreds of gigabytes of data for the first time may tr igger a need to reconsider data management options. For others, it may take tens or hundreds of terabytes before data size becomes a significant consideration." [19] Contents 1 Definition 2 Examples 2.1 Big science 2.2 Science and research 2.3 Government 2.4 Private sector 2.5 International development 3 Market 4 Architecture 5 Technologies 6 Research activities 7 Critique 7.1 Critiques of the Big Data paradigm 7.2 Critiques of Big Data execution 8 See also 9 References 10 Further reading

Definition Big data usually includes data sets with sizes beyond the ability of commonly us ed software tools to capture, curate, manage, and process the data within a tole rable elapsed time [20]. Big data sizes are a constantly moving target, as of 20 12 ranging from a few dozen terabytes to many petabytes of data in a single data set. With this difficulty, new platforms of "big data" tools are being develope d to handle various aspects of large quantities of data. In a 2001 research report[21] and related lectures, META Group (now Gartner) ana lyst Doug Laney defined data growth challenges and opportunities as being threedimensional, i.e. increasing volume (amount of data), velocity (speed of data in and out), and variety (range of data types and sources). Gartner, and now much of the industry, continue to use this "3Vs" model for describing big data.[22] I n 2012, Gartner updated its definition as follows: "Big data are high volume, hi gh velocity, and/or high variety information assets that require new forms of pr ocessing to enable enhanced decision making, insight discovery and process optim ization."[23] TBDI Definition of Big data: Big Data is a term applied to voluminous data objec ts that are variety in nature structured, unstructured or a semi-structured, inc luding sources internal or external to an organization, and generated at a high degree of velocity with an uncertainty pattern, that does not fit neatly into tr aditional, structured, relational data stores and requires strong sophisticated information ecosystem with high performance computing platform and analytical ca pabilities to capture, process, transform, discover and derive business insights and value within a reasonable elapsed time Examples Examples include Big Science, web logs, RFID, sensor networks, social networks, social data (due to the social data revolution), Internet text and documents, In ternet search indexing, call detail records, astronomy, atmospheric science, gen omics, biogeochemical, biological, and other complex and often interdisciplinary scientific research, military surveillance, forecasting drive times for new hom e buyers, medical records, photography archives, video archives, and large-scale e-commerce. Big science The Large Hadron Collider experiments represent about 150 million sensors delive ring data 40 million times per second. There are nearly 600 million collisions p er second. After filtering and refraining from recording more than 99.999% of th ese streams, there are 100 collisions of interest per second.[24][25][26] As a result, only working with less than 0.001% of the sensor stream data, t he data flow from all four LHC experiments represents 25 petabytes annual rate b efore replication (as of 2012). This becomes nearly 200 petabytes after replicat ion. If all sensor data were to be recorded in LHC, the data flow would be extrem ely hard to work with. The data flow would exceed 150 million petabytes annual r ate, or nearly 500 exabytes per day, before replication. To put the number in pe rspective, this is equivalent to 500 quintillion (51020) bytes per day, almost 20 0 times higher than all the other sources combined in the world. Science and research When the Sloan Digital Sky Survey (SDSS) began collecting astronomical data in 2000, it amassed more in its first few weeks than all data collected in the h istory of astronomy. Continuing at a rate of about 200 GB per night, SDSS has am

assed more than 140 terabytes of information. When the Large Synoptic Survey Tel escope, successor to SDSS, comes online in 2016 it is anticipated to acquire tha t amount of data every five days.[5] Decoding the human genome originally took 10 years to process; now it can be achieved in one week.[5] Computational social science Tobias Preis et al. used Google Trends data to demonstrate that Internet users from countries with a higher per capita gross do mestic product (GDP) are more likely to search for information about the future than information about the past. The findings suggest there may be a link betwee n online behaviour and real-world economic indicators.[27][28][29] The authors o f the study examined Google queries logs made by Internet users in 45 different countries in 2010 and calculated the ratio of the volume of searches for the com ing year ( 2011 ) to the volume of searches for the previous year ( 2009 ), which they c all the future orientation index .[30] They compared the future orientation index t o the per capita GDP of each country and found a strong tendency for countries i n which Google users enquire more about the future to exhibit a higher GDP. The results hint that there may potentially be a relationship between the economic s uccess of a country and the information-seeking behavior of its citizens capture d in big data. Government In 2012, the Obama administration announced the Big Data Research and Develo pment Initiative, which explored how big data could be used to address important problems facing the government.[31] The initiative was composed of 84 different big data programs spread across six departments.[32] Big data analysis played a large role in Barack Obama's successful 2012 re-e lection campaign.[33] The United States Federal Government owns six of the ten most powerful super computers in the world.[34] The NASA Center for Climate Simulation (NCCS) stores 32 petabytes of climate observations and simulations on the Discover supercomputing cluster.[35] The Utah Data Center is a data center currently being constructed by the Uni ted States National Security Agency. When finished, the facility will be able to handle yottabytes information collected by the NSA over the Internet.[36][37] Private sector Amazon.com handles millions of back-end operations every day, as well as que ries from more than half a million third-party sellers. The core technology that keeps Amazon running is Linux-based and as of 2005 they had the world s three lar gest Linux databases, with capacities of 7.8 TB, 18.5 TB, and 24.7 TB.[38] Walmart handles more than 1 million customer transactions every hour, which is imported into databases estimated to contain more than 2.5 petabytes (2560 te rabytes) of data the equivalent of 167 times the information contained in all th e books in the US Library of Congress.[5] Facebook handles 50 billion photos from its user base. FICO Falcon Credit Card Fraud Detection System protects 2.1 billion active a ccounts world-wide.[39] The volume of business data worldwide, across all companies, doubles every 1 .2 years, according to estimates.[40] Infosys has also launched the BigDataEdge to analyse the Big data.[41][42] Windermere Real Estate uses anonymous GPS signals from nearly 100 million dr ivers to help new home buyers determine their typical drive times to and from wo rk throughout various times of the day. [43] International development Following decades of work in the area of the effective usage of information and communication technologies for development (or ICT4D), it has been suggested tha

t Big Data can make important contributions to international development.[44][45 ] On the one hand, the advent of Big Data delivers the cost-effective prospect t o improve decision-making in critical development areas such as health care, emp loyment, economic productivity, crime and security, and natural disaster and res ource management.[46] On the other hand, all the well-known concerns of the Big Data debate, such as privacy, interoperability challenges, and the almighty powe r of imperfect algorithms, are aggravated in developing countries by long-standi ng development challenges like lacking technological infrastructure and economic and human resource scarcity. "This has the potential to result in a new kind of digital divide: a divide in data-based intelligence to inform decision-making." [46] Market "Big data" has increased the demand of information management specialists in tha t Software AG, Oracle Corporation, IBM, Microsoft, SAP, EMC, and HP have spent m ore than $15 billion on software firms only specializing in data management and analytics. In 2010, this industry on its own was worth more than $100 billion an d was growing at almost 10 percent a year: about twice as fast as the software b usiness as a whole.[5] Developed economies make increasing use of data-intensive technologies. There ar e 4.6 billion mobile-phone subscriptions worldwide and there are between 1 billi on and 2 billion people accessing the internet.[5] Between 1990 and 2005, more t han 1 billion people worldwide entered the middle class which means more and mor e people who gain money will become more literate which in turn leads to informa tion growth. The world's effective capacity to exchange information through tele communication networks was 281 petabytes in 1986, 471 petabytes in 1993, 2.2 exa bytes in 2000, 65 exabytes in 2007[15] and it is predicted that the amount of tr affic flowing over the internet will reach 667 exabytes annually by 2013.[5] Architecture Due to the complexity of big data systems, a sophisticated practice of big data architecture is a must. The Big Data Architecture Framework (BDAF) is an archite cture framework for Big Data solutions, aimed at helping manage a set of discret e artifacts and implementing a collection of specific design elements.[47] The p urpose of BDAF is to enforce the adherence to a consistent design approach, redu ce system complexities, enhance loose-coupling, maximize reuse, decrease depende ncies, and increase productivity. BDAF comprises four integral parts: Domain-specific, enablement-dependent, platf orm-agnostic, and technology-neutral model. The BDAF components are model-centri c and architecture-driven, forming a cohesive construct for big data processing, including data extract, ingestion, store, processing, schema, aggregation, mess aging, interfacing, reporting, visualization, monitoring, streaming, and automat ion. In 2004, Google published a paper on a process called MapReduce that used such a n architecture. MapReduce framework provides a parallel programming model and as sociated implementation to process huge amount of data. With MapReduce, queries are split and distributed across parallel nodes and processed in parallel (the M ap step). The results are then gathered and delivered (the Reduce step). The fra mework was incredibly successful,[clarification needed] so others wanted to repl icate the algorithm. Therefore, an implementation of MapReduce framework was ado pted by as an Apache open source project named Hadoop.[48] MIKE2.0 is an open approach to information management. The methodology addresses handling big data in terms of useful permutations of data sources, complexity i n interrelationships, and difficulty in deleting (or modifying) individual recor ds.[49] Technologies

File:DARPA s Topological Data Analysis program seeks the fundamental structure of massive data sets and is developing the tools to exploit that knowledge.tiff DARPA s Topological Data Analysis program seeks the fundamental structure of massi ve data sets. Big data requires exceptional technologies to efficiently process large quantiti es of data within tolerable elapsed times. A 2011 McKinsey report[50] suggests s uitable technologies include A/B testing, association rule learning, classificat ion, cluster analysis, crowdsourcing, data fusion and integration, ensemble lear ning, genetic algorithms, machine learning, natural language processing, neural networks, pattern recognition, anomaly detection, predictive modelling, regressi on, sentiment analysis, signal processing, supervised and unsupervised learning, simulation, time series analysis and visualisation. Multidimensional big data c an also be represented as tensors, which can be more efficiently handled by tens or-based computation,[51] such as multilinear subspace learning.[52] Additional technologies being applied to big data include massively parallel-processing (MP P) databases, search-based applications, data-mining grids, distributed file sys tems, distributed databases, cloud based infrastructure (applications, storage a nd computing resources) and the Internet.[citation needed] Some but not all MPP relational databases have the ability to store and manage p etabytes of data. Implicit is the ability to load, monitor, back up, and optimiz e the use of the large data tables in the RDBMS.[53] DARPA s Topological Data Analysis program seeks the fundamental structure of massi ve data sets and in 2008 the technology went public with the launch of a company called Ayasdi. The practitioners of big data analytics processes are generally hostile to slowe r shared storage[54], preferring direct-attached storage (DAS) in its various fo rms from solid state disk (SSD) to high capacity SATA disk buried inside paralle l processing nodes. The perception of shared storage architectures SAN and NAS is th at they are relatively slow, complex, and expensive. These qualities are not con sistent with big data analytics systems that thrive on system performance, commo dity infrastructure, and low cost. Real or near-real time information delivery is one of the defining characteristi cs of big data analytics. Latency is therefore avoided whenever and wherever pos sible. Data in memory is good data on spinning disk at the other end of a FC SAN c onnection is not. The cost of a SAN at the scale needed for analytics applicatio ns is very much higher than other storage techniques. There are advantages as well as disadvantages to shared storage in big data anal ytics, but big data analytics practitioners as of 2011 did not favour it.[55] Research activities In March 2012, The White House announced a national "Big Data Initiative" that c onsisted of six Federal departments and agencies committing more than $200 milli on to big data research projects.[56] The initiative included a National Science Foundation "Expeditions in Computing" grant of $10 million over 5 years to the AMPLab[57] at the University of Califo rnia, Berkeley.[58] The AMPLab also received funds from DARPA, and over a dozen industrial sponsors and uses big data to attack a wide range of problems from pr edicting traffic congestion[59] to fighting cancer.[60] The White House Big Data Initiative also included a commitment by the Department of Energy to provide $25 million in funding over 5 years to establish the Scala ble Data Management, Analysis and Visualization (SDAV) Institute,[61] led by the Energy Department s Lawrence Berkeley National Laboratory. The SDAV Institute aim

s to bring together the expertise of six national laboratories and seven univers ities to develop new tools to help scientists manage and visualize data on the D epartment s supercomputers. The U.S. state of Massachusetts announced the Massachusetts Big Data Initiative in May 2012, which provides funding from the state government and private compan ies to a variety of research institutions.[62] The Massachusetts Institute of Te chnology hosts the Intel Science and Technology Center for Big Data in the MIT C omputer Science and Artificial Intelligence Laboratory, combining government, co rporate, and institutional funding and research efforts.[63] The European Commission is funding a 2-year-long Big Data Public Private Forum t hrough their Seventh Framework Program to engage companies, academics and other stakeholders in discussing Big Data issues. The project aims to define a strateg y in terms of research and innovation to guide supporting actions from the Europ ean Commission in the successful implementation of the Big Data economy. Outcome s of this project will be used as input for Horizon 2020, their next framework p rogram.[64] Critique Critiques of the Big Data paradigm come in two flavors, those that question the implications of the approach itself, and those that question the way it is curre ntly done. Critiques of the Big Data paradigm "A crucial problem is that we do not know much about the underlying empirical mi cro-processes that lead to the emergence of the[se] typical network characterist ics of Big Data" [65]. In their critique, Snijders, Matzat, and Reips point out that often very strong assumptions are made about mathematical properties that m ay not at all reflect what is really going on at the level of micro-processes. M ark Graham has leveled broad critiques at Chris Anderson's assertion that big da ta will spell the end of theory: focusing in particular on the notion that big d ata will always need to be contextualized in their social, economic and politica l contexts.[66] Even as companies invest eight- and nine-figure sums to derive i nsight from information streaming in from suppliers and customers, less than 40% of employees have sufficiently mature processes and skills to do so. To overcom e this insight deficit, "big data", no matter how comprehensive or well analyzed , needs to be complemented by "big judgment", according to an article in the Har vard Business Review.[67] Much in the same line, it has been pointed out that the decisions based on the a nalysis of Big Data are inevitably "informed by the world as it was in the past, or, at best, as it currently is".[46] Fed by a large number of data on past exp eriences, algorithms can predict future development if the future is similar to the past. If the systems dynamics of the future change, the past can say little about the future. For this, it would be necessary to have a thorough understandi ng of the systems dynamic, which implies theory.[68] As a response to this criti que it has been suggested to combine Big Data approaches with computer simulatio ns, such as agent-based models, for example.[46] Those are increasingly getting better in predicting the outcome of social complexities of even unknown future s cenarios through computer simulations that are based on a collection of mutually interdependent algorithms. In addition, use of multivariate methods that probe for the latent structure of the data, such as Factor Analysis and Cluster Analys is, have proven useful as analytic approaches that go well beyond the bi-variate approaches (cross-tabs) typically employed with smaller data sets. Consumer privacy advocates are concerned about the threat to privacy represented by increasing storage and integration of personally identifiable information; e xpert panels have released various policy recommendations to conform practice to expectations of privacy.[69]

Critiques of Big Data execution Danah Boyd has raised concerns about the use of big data in science neglecting p rinciples such as choosing a representative sample by being too concerned about actually handling the huge amounts of data.[70] This approach may lead to result s bias in one way or another. Integration across heterogeneous data resources so me that might be considered "big data" and others not presents formidable logist ical as well as analytical challenges, but many researchers argue that such inte grations are likely to represent the most promising new frontiers in science.[71 ]