The Mechanics of Crowdsourcing: Mobilizing the Many and the Extraordinary

Video in TIB AV-Portal: The Mechanics of Crowdsourcing: Mobilizing the Many and the Extraordinary

Formal Metadata

The Mechanics of Crowdsourcing: Mobilizing the Many and the Extraordinary
Title of Series
Number of Parts
CC Attribution - ShareAlike 3.0 Germany:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this license.
Release Date

Content Metadata

Subject Area
Crowdsourcing is an integral part of the digital society. Crowdsourcing still seems to carry significant novelty for scholars and practitioners alike, but, arguably, it has become a buzzword. As a result, the meaning of crowdsourcing is blurred and distorted. In our talk we reassess and clarify the concept "crowdsourcing". Furthermore, we distinguish between four types of crowdsourcing and examine the consequences of each type for (a digital) society.
Point (geometry) Group action Decision theory Orientation (vector space) Mass Power (physics) Broadcasting (networking) Mechanism design Goodness of fit Hypermedia Internetworking Different (Kate Ryan album) Forest Data acquisition Authorization Selectivity (electronic) Extension (kinesiology) Position operator Task (computing) Physical system Area Addition Focus (optics) Instance (computer science) Auto mechanic Message passing Arithmetic mean Personal digital assistant Telecommunication Network topology Self-organization Quicksort Family
Axiom of choice Context awareness Web crawler Euler angles Multiplication sign View (database) Modal logic Source code Set (mathematics) Client (computing) Open set Likelihood function Neuroinformatik Word Sign (mathematics) Mechanism design Type theory Bit rate Radio-frequency identification Different (Kate Ryan album) Forest Similitude (model) Finitary relation Information Office suite Endliche Modelltheorie Information security Dissipation Source code Channel capacity Mapping Moment (mathematics) Sound effect Instance (computer science) Web crawler Open set Type theory Data mining Process (computing) Sample (statistics) Order (biology) Chain Self-organization Summierbarkeit Metric system Web page Classical physics Point (geometry) Onlinecommunity Dataflow Game controller Service (economics) Identifiability Link (knot theory) Channel capacity Similarity (geometry) Online help Number Product (business) Wave packet Power (physics) Twitter Software testing Angular resolution Associative property Metropolitan area network Computing platform Task (computing) Mobile Web Home page Standard deviation Information Compass (drafting) Forcing (mathematics) Variance Planning Volume (thermodynamics) Software maintenance Power (physics) Similarity (geometry) Performance appraisal Mathematics Personal digital assistant Hydraulic jump Family Spectrum (functional analysis)
Kepler conjecture Presentation of a group Game controller Divisor Line (geometry) Multiplication sign Channel capacity Control flow Frustration Mereology Code Rule of inference Power (physics) Traffic shaping Expected value Mechanism design Type theory Radio-frequency identification Forest Core dump Finitary relation Species Information Process (computing) Task (computing) Shift operator Distribution (mathematics) Theory of relativity Physical law Auto mechanic Exploit (computer security) Power (physics) Vector potential Network topology Order (biology) Video game Right angle Bounded variation Family Thermal conductivity Astrophysics
Data acquisition
the the
it not
good good afternoon everybody things for joining us here I guess with all the to 3 more days threatened because we're getting toward the final so on behalf of all of almost and myself by I wish all of us interesting uh and so if the system has the authority has found and have found is something that may sound familiar to soften orientation have you heard of this have had relationships 1 2 uh I would have had to have 1 feature that the individual or an organization and this agent is capable of reaching everybody else so for instance of our agent can or individual or organization can broadcast a message and everybody else will receive the message more all be on the receiving end everybody makes that decision how they want to react to that of message and then uh that could include getting back to the central agent of he can orient to take any action or whatever and the traditionally 1 to all relationships of the associated with a lot of power and were thus in the center of political focus think of mass media which additional channels of 4 1 all communication think afraid UT-VE newspaper it's all about 1 all communication and traditionally extensive use of mass media was reserved to the power of and now we all know when the internet came it came with a promise and a promise was each and everybody of us will be capable of engaging in 1 to all communication the that's quite a powerful promise because it means that everyone on lion is in principle capable of reaching everybody else online and by now that means reaching about 80 per cent of the developed world and about 1 3rd of the developing world and in addition we know about of the backbone of the Internet of the global Internet use cell phone infrastructure and the so for infrastructure is by now completed no investment goes into upgrading in infrastructure so that it can deliver broadband Internet so it will not be very long until fact technically where in principle in a situation where every individual on this planet is in a position that they can reach every 1 else on this planet and in fact even now the Internet is delivered on many of these promises us think of the many marvelous examples that you've seen here at the point precursor on how individuals or small initiatives managed to mobilize fairly big crowds of for political purposes or another important area how many great examples have we seen all of entertaining joyful material that because of the Internet can now reach the crowds and reach big audiences so in the realm of social activism of politics of entertainment that 1 to all communication is actually has become standard crowdsourcing also uses the 1 all communication but it raises a completely different question because in politics and in entertainment the bigger your audience the better but crowdsourcing is not primarily about politics or entertainment it's about solving problems it's about executing tasks that means now we have economic criteria to fulfill that means you wanna solve your problem with the least possible effort with using this little resources as possible and the same is true for executing tasks you want to execute them as efficiently as possible and that raises the question why would you want to go for a 1 to all communication and talk to everybody usually you would want to that you want to to a few people that actually help you sort of have had their hair case how makes sense to address everybody and go from 1 to all communications on a footnote it's often said that uh crowdsourcing means you outsource a task to across that's somewhat imprecise it's imprecise because in many cases is true that the crowd actually solves the task but there are many other cases in which the crowd is only more realize this intermediary step and you want to identify 1 or a few people to who you actually all sorts of tasks so to be more specific of working definition of crowdsourcing is you use broadcast search you use logically speaking global search to connect to outsourcing partners for problem solving and task execution now initiatives crowdsourcing initiatives are now in the hundreds and were in a situation where it's getting difficult to see the forest for the trees all a claim is that if you carefully look at how the US has one to all communication actually generate value for problem solving and task execution if you carefully pursue this question you find out that the underlying these hundreds of initiatives are the basic mechanisms that help us better understand the crowdsourcing and this is what wrapped in this talk reassessing and clarifying the concept and our guidelines for doing this is pursuing the question how does global search actually create value now the idea of the global search creates value has nothing to do with the Internet its way over so let's look at 2 examples that date back to linear the Chinese legends have it's that and you've just fix of century after christ the emperor wanted to expand his empire and that's needed a the army so how did he assemble that big army which soldiers should you pick up he came up with a very simple solution he asked every family you have to conclude and 1 media because this task is so humungous so be that selection wouldn't be efficient he went for everyone so that
everyone in his community everyone in his target for every family was supposed to contribute 1 male and it wouldn't matter if the individual contribution was excellent or mediocre and number tissue volume of contributions would actually help them solve the task now for a different what seemed similarly old but a very different approach another story uh I'm not sure if you're particularly religious crowd this is a baby cries being born and those of you who know the Bible no that wouldn't be the Christ was born the reduced so wise man is the guys not celebrities that came to warships the baby but they didn't know where to go so they went to the local king and the local came where should we go we wanna pay attention and out on the new king as you can imagine the acting King wasn't induced about the idea of thinking being born and decided to get rid of the new king yet he had no clue who to get rid of it so what should you do 1 to also he decided to get rid of every new boy because not because he didn't like every new newborn it was just 1 guy that he was after very selective case no but in order to get to this 1 getting to everybody would be an excellent solution so we have 2 very different fundamental principles 1 is very big tasks 1 is very selective tasks and we go deeper into these 2 principles and fewer I really less leading to rescue effect friends 1st century I hand as 1 of mentioned the principles remain the same and we want to expand and messier than more in detail how to play out today so we distinguish between 4 types of value and 2 types of associate with a selective approach so mobilize many to find 1 solution I want another 1 is an anchored aggregated approach so mobilize many to harness the all the contribution exploit the sheer force of numbers I will illustrate those types of them in examples you probably know some or most of the examples of ill rush through them and just for illustration so the first one is creative expertise I'm looking for a new and valuable contribution that requires specific skills and the knowledge for this is not in that organization or we don't know where to find it and the more people we reach the more the higher the likelihood is that we find a good solution most known example we heard also today is InnoCentive and crowdsourcing contest that form the organizations can pass the crowd to solve an innovation platform the crowd metrics submit solutions and the client the organization on the side decides who will and so in this case in this and that is an intermediary in the middle who connects organizations once a solvent innovation problems the crowd another example from Berlin takes but different approach a more community-driven and and it's a fair approach they also do this kind of innovation contest but this seperate 1st the community decides who wins so there's amount of money a prize Richie Millicent's who's winning and who's not winning and 2nd the client needs to pay a 2nd time so there's so it's variance and softly clients speed 2 times for the community prize for the Committee decides and I our four-day idea the client wants to use another approach is critical information critical for information is the search for context-specific personal experience and knowledge and in this difficult than our creative expertise but for for critical information you don't need domain-specific training it's more about being at the right place at the right moment and now that's the only time I go back to history again as the wanted posters the perfect example for that I know you're Davila because you're standing next to me and I can identify you and call the police so that's a classical example of critical information another 1 example are used user communities like disparate communities like office 3 6 5 this is not an advertisement and where it big firms outsource the customer services to get IVA problem the product Robert knows the answer because it is similar problem with that because his personal experience he could answer of my problem the 3rd type of value so maybe I should we give so the first one was we need to find someone who has a great idea 2nd was critical information we need to find someone who has a personal experience those who the villainous and now we turn to attitude types produce rebalanced B 1 ask everyone we wanna use B 1 mobilize as many as possible and make use of the numbers and here we go the 4th type of 3rd type of for creating value is execution capacity so this is about the search for a large number of small contributions that add up and solve a big task and a mostly task which cannot yet be solved by computers again known examples most people have filled about recapture for you this is a security feature on that page you need to type in those numbers and by typing in those numbers you help Google to improve other map services from the public realm moons soon soon URIs you the crowd you help those guys execution opacity very small standardized task help to identify moon craters you just go there and present that's where readers a greater and also
crowdfunding relies on a similar principle research for a large number of users were willing to delay the small small amount of money but all if it adds up it solves a big a big pot of cash the forest type is would be called bargaining power so this is about the search for a large number of similar offerings to lower prices because crowd mobilization if I mobilize many this trades abundance never have abundance this can be turned to economic gains for those guys who to the crowd sourcing and intermediates and examples here Bob Wills of unknown 99 design they have assigned condescend the mobile platform over all of 40 thousand templates because of all the choices it's easy to the people of the link to deliver for low prices another 1 is the lens which is technically and on freelance market for you posted job you get many proposals because you get so many proposals it's all transparent it's very likely Devereux always be someone whom the officer lower price when you put a poster dropped and of course those 4 types do not always have cure indep you form it's always overlapping take Amazon Mechanical Turk which is a marketplace for tasks where people go online and unsolved us like categorizing home pages and execution capacity in this case of course give many small tasks we mobilize a large number of users and this is a very efficient all adds up to solving a big problem bargaining power sense because we mobilize to many users the people use Mechanical Turk are able to pay below minimum wage and very low payments because so many people there's always someone else to do the job another example Marie of combined of course is a cyclic fixed which is something to fit that to fix potholes and or a maintenance problems and use the critical information again I need to be in trial on down the street to see that there is a pothole put on my on my mobile phone take a picture of the poll took all posted on the platform so try nose OK we need to fix this so this is the critical information by the execution capacity pardon the spectrum is by outsourcing the maintenance work to the crowd the of 24 7 surveillance and quality control of the streets of Toronto because the crawlers everywhere and can never take pictures it would make no sense which to do this task itself so these are the floor types which are included in any crowdsourcing I had from the trade value but but I read the sources 1 of the societal sources that this actually works yeah and we argue crowdsourcing mainly mobilize from the fringes of a global economy so the first one which is actually from a German business book I like this quote from Christophe of geezer in the initial it allows and I'm interested in English people participate in open source and crowdsourcing activities in the free time to make up for the lack of excitement in their daily jobs of someone motivations of course other people wanna do interesting things other motivations are well known like I wanna learn and especially if I'm in the early stage of Korea is many people do crowdsourcing because the 1 the signal the skills that won a bill that the repetition which is essentially a very important in an attention economy yeah and not a source of value is of you like to call of marginal or the effect of killing time so you doing for example microtasks or small rating test on the subway and as he painted perfectly noted is the less investment necessary for the intervals but dissipate because the task is just 1 click the higher the like a set of contributors and the 3rd source it is yeah for mobilizing the people is especially in emerging economies so platforms many crowdsourcing of engineers up to exploit the fact that burning small sums via microtasking or reading a compass just once in a while just once in a while even if the bringing chances are so slim is still economically feasible with an emerging countries an example that reflects so this data from Ross from 2010 there you can see the North American population on Mechanical Turk decreasing at the same time India is increasing in here if the turning point 2010 we don't have a better data but we have also added anecdotal evidence of this trend is continuing for example here we have designed mining and assigns the design competition model works especially good in Asia and the sign crowd transcends Asia they go up abscissa is that the mobilize the crowd in Asia and getting more clients and because you had the conference and the conference is always stands for what the industry very industries heading and here the 1st crowdsourcing global in Singapore and all the next year in singer person we see the crowdsourced of the crowds of population all the shifting Asia so we have another 5 minutes I suppose that awareness that leave us so we have these 4 types of value that is created through crowdsourcing and now we could go into much detail and look what are the specific challenges for each of these types we don't have the time you wouldn't we did not plan on doing this just give you 1 example but before that note that we have 2 different types of challenges 1 is the the mechanics of how does it actually work and this is pretty much about how best information flow from 1 to all and back so we have to make sure we have an uninterrupted communicate communicative chain from 1 to all and back and that's not always easy so 1 example creative expertise we say it's difficult to to communicate the problems and evaluate solutions why would that be the case clearly if you have a problem that you cannot solve and you don't know how to talk to it's difficult to describe the problem in a way that is both meaningful and doesn't narrowed down the problem to where it's unsolvable because it's a limited view of the problem so for instance when nozzle tried to win as outsourced of crowdsourced predicting solar activity a problem that they hadn't been able to solve in years despite a lot of investment they had to defined the problem in a way that was
different was broad enough so that everybody could understand it and in fact it was a retired of telecommunications engineer not an astrophysicist that actually solved the same is true for evaluating the problem if you're far from your own expertise difficulty value but the more important part here is the other side the this has to do with the societal challenges this has to do with shaping social relations and here are important challenges in crowdsourcing few examples winner-take-all competition uh Thomas mentioned and you will to allow goes into fostering the community because we know from InnoCentive that it can be quite frustrating for contributors if you know you have at best law that we like chances to win because only 1 will win out of many many contributions so there's a potential for frustration there in order to make it fair and sustainable of 1 approach is close to the community our but there are also interesting implications about communities because think about this big data peak rather we watched the streets of Toronto we watched the streets of New York or Trip Advisor that we watch all hotels what enormous amount of control those that make possible no hotel company would be able to control their employees only close to about all the customers can do so it's not only about getting more community getting more crowd it's also about developing crowds competence developing codes of conduct how strong crowd can actually the develop that added value without doing damage the damage here the lynch mob you remember maybe Boston Marathon some people thought they saw the perpetrator posted pictures somebody was wrongfully accused and they're almost destroyed his life so death interesting aspects in here including all problems exploitation we have seen this at the beginning of industrialization simple tasks very little of very low wages expectations so we unionized developed crowds competence to re shift of the income distribution so it's this part that we think is the future of crowdsourcing their challenges here but they are relatively close to being solved this is the interesting part and this has to do with developing more powerful crops and more powerful crawls are good because they can resist but they're also they have a life of their own so if we go back to assembling an army we should also keep in mind but this is the army that was assembled by the East Indian company and the East India Company the Brits had no idea about the local communities and the rules and the culture so we without surprised to hear that it didn't take long until the crowd turned against those who actually had recruited them so to summarize crowdsourcing is proliferating and multiplying and we're going to see many more variations so it will be difficult to see the forest for the trees however the in a simple it's 4 mechanisms and the things that we were able to explain this basic 4 mechanisms 2nd where does the value actually come from it comes from the fringes marginalized resources are used through crowdsourcing are connected to the core of the economy which has yeah pros and cons finally where's the future of the crowds at getting more political a Kickstarter is being sent flora of fraudulent use of crowd funds of CrowdFlower use some for minimum wage so there's a lot of movement there which is important on the other hand we know it's not only about making crawled strong it's also about developing community competence to avoid big brother phenomena or a lynch mobs in the ditch which I think this is where we can stop and hopefully we can take a few more questions thank you very much for the presentation I think that's within because thank ifyou time's running up a thing we only have time for 1 or 2 questions but some a with the question not did you any questions boolean that at the at the at the the at the you're much well then I I have a question for you um it's it's really interesting that you that you mentioned in the end there could be some kind of an immense supertree potential in the crowd so the crowd fights spec plants and what kind of factors what kind of the arrivals must be there in order for crowds to buy take to become resilient that that's interesting I think that the 2 factors 1 this of crowds like getting more self-confident they understand how much value they contribute and that's the receive their own right to ask the other thing is that I think intermediaries a really important and the most advanced into the intermediaries and the thinking of correctly and you will look quite interesting examples they actually helped to crawl to emancipate because they understand if the crowd is stronger it not only gains more value it also contributes more about so there's a win-win situation also in there as long as the political power and the Community competence of both that reminds us with you it OK I think because of our time and thanks again for your presentation and after a short break we continue with our next 1 we think you're and