Computational Thinking - Учебное пособие (1176923), страница 18
Текст из файла (страница 18)
Usingonline social networks for healthcare provides an opportunity toanalyze behavioral data and study social structures formed as aresult of ties to health behaviors. Both the American CancerSociety and the Centers for Disease Control and Prevention, forexample, have experimented with virtual communities such asSecond Life to test whether social multimedia can help spread100the word about such issues as nutrition awareness, cancerscreening, and infectious-disease prevention.Collaboratively monitoring health status across media withubiquitous wireless sensors and cameras could also helpmedical staff craft an effective healthcare application. Forexample, an ECG sensor carried by an elderly person withphysical disabilities might capture any unusual heartbeat ratesand send this information through a social network of familymembers, physicians, friends, and emergency services; theinformation's context — at home or outdoors, for example —could help physicians make time-critical decisions.
However,the usage of such sensor data will also give rise to privacy issues.In practice, it could be less privacy invasive if these sensor dataare strictly restricted for use in healthcare applications.Security applicationsWe can characterize many security applications as socialmultimedia computing applications. They've emergedespecially rapidly in recent years with the proliferation of videosurveillance systems in a wide variety of domains, such ashomes, banks, airports, and convenience stores.
Variousgovernment entities have used social network analysis to analyzeterrorist networks, communications, criminal organizations, andresources. However, to date, few works have focused on socialnetwork analysis in surveillance video.In spite of the possible invasion of privacy, mastering the rolesuch networks play in monitoring surveillance video data is ofgreat interest to law enforcement and homeland security. Forexample, DARPA is soliciting innovative research proposals todevelop the Persistent Stare Exploitation and Analysis Systemfor automatically and interactively discovering actionableintelligence through wide-area-threat analysis of complexmotion imagery surveillance of urban, suburban, and ruralenvironments.Although work in social network analysis, multimedia contentanalysis, and other disciplines can be adapted to socialmultimedia computing systems, numerous problems inmodeling, analyzing, and utilizing multimedia social networks101have yet to be solved.Multimedia social dynamicsSocial networks often involve many users of different types —from rational to selfish and malicious — all with differentobjectives.
Thus, modeling and analyzing user behaviors andsocial dynamics poses a fundamental challenge to helpstimulate user cooperation, maximize overall systemperformance, and minimize the damage caused by malicioususers.Several specific multimedia properties make analyzingmultimedia social dynamics different from traditional socialnetwork analysis. First, user behaviors are highly dynamic,especially when users watch live streaming video on the samewireless network or share the same limited backboneconnection to the Internet.Second, the modeling and analysis of user behaviors aremostly content-relevant. In colluder social networks, forexample, multimedia fingerprinting or other contentidentification technologies could model user behavior and trackpeople who illegally use copyrighted multimedia.Third, the potential rewards are time-sensitive.
For example,the earlier a colluded copy is released, the more people will bewilling to pay for it. Thus, all colluders have an incentive tomount collusion as soon as possible.Multimedia interaction dynamicsIntuitively, developing new multimedia social interaction toolsrequires a close integration of sociology, multimedia, andcommunications technologies. Specifically, the social sciencestudies addressing the dynamics of large-scale socialinteraction and activity might revitalize research into novelsocial interaction methods and tools. Twitter, originallyconceived as a mobile status update service that provides aneasy way to keep in touch with friends, offers one example.Twitter users send and receive short, frequent answers to onequestion, "What are you doing?"However, Twitter changed that question to "What's happening?"so that people, organizations, and businesses could leverage the102network's open nature to share anything they wanted, includingpictures and video.
This shift created a new kind of informationnetwork in its users' social space. In this sense, social scientistsshould take a more active role in coping with the challenge ofdeveloping new multimedia tools to enable more powerful socialinteraction.A greater challenge requires developing user interfaces andinteraction paradigms to allow seamless communication andinteraction with remote and virtual environments. In practice,this vision is achievable by combining new sensors that covertouch, smell, taste, and motion; immersive output devices suchas large displays, and 3D technology.Social multimedia community analysisSocial multimedia provides several effective ways to harvest thelarge-scale digital traces of social behaviors, such as onlinecontent sharing, vlogging, and video surveillance. With theincreasing availability of such pervasive data, key researchchallenges will involve developing methodologies for largescale validations of social science theories and for new theoriesand inferential analysis methods that can analyze this kind ofnew data.For example, several studies have focused on social behaviorand organization dynamics in online text-blog communities.Without considering the multimedia properties of thecommunity structure, these social network analysis models canbe directly used in vlog communities.
However, the problembecomes much more complex if we consider the interrelationamong the content, social, and temporal dimensions of vlogs.For example, large-scale experiments are needed to verifywhether the "six-degrees-of-separation principle" remains validin vlog communities, given that the change of media formsfrom text-only to video might speed up information propagationand consequently shorten networks.
This study could shed newinsights into real-world applications such as online advertisingand viral marketing.Privacy protectionMost social multimedia data are proprietary, such as user103profile data — name, place, date of birth, and e-mail address— or ancillary data such as the IP address or time ofconnection. Properly managing privacy issues is essential bothto facilitate research and safeguard consumer privacy.Nevertheless, the privacy issues surrounding social multimediadata are more complex. Generally speaking, users should havethe right to control their personal data, which implies gettingaccess to the data, modifying it, asking for corrections, orasking for deletion.
However, this right isn't easily guaranteedin the context of social multimedia. For example, users areusually eager to share pictures in online communities, butthese images can easily be used for secondary purposes suchas face recognition and image retrieval, especially when taggedwith metadata such as name, e-mail address, and physicaladdress of the person pictured. Because a single dramaticincident involving a breach of privacy could produce rules andstatutes that stifle the nascent field of social multimediacomputing, a systematic study of privacy issues and theircorresponding technological, procedural, and rule-relateddevelopments must be undertaken to reduce security risks andpreserve research potential.
The explosive growth of socialmultimedia on the Internet is revolutionizing the waycontent distribution and social interaction work whilepresenting an evolving multidisciplinary research andapplication field. At present, this research addresses thedescriptive analysis level, but the potential for developingsocial multimedia computing theories and methods remainspromising.Notes:A milestone - a stone set up beside a road to mark thedistance in miles to a particular place; figuratively - an actionor event marking a significant change or stage in developmentA vlog – a video blog or video log (pronounced 'vlog' or 'vlog'), a form of blog for which the medium is videoA P2P network – a peer-to-peer; denoting a network or datacommunications in which no dedicated server is involved104NAT - Network address translationFree-riding - enjoying a service without paying for itECG (or EKG from Greek: kardia, meaning heart) ElectrocardiographyExercises:1.
Give Russian equivalents to the following words andphrases and explain them in your own words:the ubiquitous presence of smth; to pose a challenge for smth;video surveillance; capture social activity and interaction; toharvest large-scale digital traces; a milestone; to harness thewisdom of crowds; a core role; live streaming video; timesensitive2. Complete the word formation using prefixes and suffixes:Interact; consumption; respond; behavioral; organize; prevail;cooperation; intrude; benefit3. Translate the sentences into Russian paying attention to thecompounds:1.
YouTube's success triggered other online content-sharingsites such as Flickr, Digg, and Youku.2. As a surrogate to content-based searches currently in theirinfancy, collaborative recommendation is an important toolfor finding multimedia content.3. They also provide better social-interaction data to facilitateanalyzing temporal interaction dynamics because entities areoften archived in reverse chronological order.4. Ideas and technologies from social multimediacomputing have recently found their applications incommunication channels such as peer-to-peer networks.5. The proliferation of user-generated content (UGC) and theassociated metadata on social multimedia sites introduces newchallenges in search, including vulnerability to spam and noise,and short lifespan.6.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
