JenniferGolbeck_2013X[詹尼弗.高尔拜克][炸薯条之迷你以为社交媒体的“赞”就是单纯的“赞”而已吗].pdf

《JenniferGolbeck_2013X[詹尼弗.高尔拜克][炸薯条之迷你以为社交媒体的“赞”就是单纯的“赞”而已吗].pdf》由会员分享，可在线阅读，更多相关《JenniferGolbeck_2013X[詹尼弗.高尔拜克][炸薯条之迷你以为社交媒体的“赞”就是单纯的“赞”而已吗].pdf（5页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、www.XiYuS锡育软件If you remember that first decade of the web,it was really astatic place.如果你还记得网络时代的头十年,网络是一个水尽鹅飞的地方。00:12You could go online,you could look at pages,and they wereput up either by organizations who had teams to do it or byindividuals who were really tech-savvy for the time.你可以上网,你可以浏览网页

2、,当时的网站 要么是由某个组织的专门团队建立,要么就是由真正的技术行家所做,这就是当时情况。00:16And with the rise of social media and social networks in theearly 2000s,the web was completely changed to a placewhere now the vast majority of content we interact with isput up by average users,either in YouTube videos or blogposts or product revie

3、ws or social media postings.但在二十一世纪初 随着社交媒体以及社交网络的兴起,网络发生了翻天覆地的变化:如今网络上大部分的互动内容 都是由大众网络用户提供,既有Youtube视频,也有博客文章,既有产品评论,也有社交媒体发布。00:27And its also become a much more interactive place,wherepeople are interacting with others,theyre commenting,theyre sharing,theyre not just reading.与此同时,互联网成为了一个有更多互动的地方

4、,人们在这里互相交流、互相评论、互相分享,而不只是阅读信息。00:46So Facebook is not the only place you can do this,but its thebiggest,and it serves to illustrate the numbers.面谱网不是唯一一个你可以做这些事情的地方,但它确实是最大的一个,并且它用数字来证明这点。00:54Facebook has 1.2 billion users per month.面谱网每个月有12亿用户。00:59So half the Earths Internet population is using

5、Facebook.由此可见,地球上一半的互联网用户 都在使用面谱网。01:02reviews:n.综述;评论;复习(review的复数);复核/v.回顾;复习(review的三单形式);检验 interacting:n.相互作用;相互制约/v.互相影响;互相作用(interact的ing形式)commenting:评论/注解/说明(comment的现在分词)They are a site,along with others,that has allowed people tocreate an online persona with very little technical skill,an

6、dpeople responded by putting huge amounts of personaldata online.这些都是网站,允许人们在网上创建不同的角色,但这些人又不需要有多少计算机技能,而人们的反应是01:06So the result is that we have behavioral,preference,demographic data for hundreds of millions of people,whichis unprecedented in history.结果是,我们拥有数以亿计人的 以及人口数据资料。这在历史上前所未有。01:17And as a

7、 computer scientist,what this means is that Ive beenable to build models that can predict all sorts of hiddenattributes for all of you that you dont even know youresharing information about.对于作为计算机科学家的我来说,这意味着 我能够建立模型 来预测各种各样的 你或许完全没有意识到的 与你所分享的信息相关的隐藏信息。01:26As scientists,we use that to help the wa

8、y people interactonline,but theres less altruistic applications,and theres aproblem in that users dont really understand thesetechniques and how they work,and even if they did,theydont have a lot of control over it.作为科学家,我们利用这些信息 来帮助人们在网上交流。但也有人用此来谋取自己的私欲,而问题是,用户并没有真正理解 其中用到的技术和技术的应用方式。即便理解了,也不见得他们有

9、话事权。01:37So what I want to talk to you about today is some of thesethings that were able to do,and then give us some ideas ofhow we might go forward to move some control back intothe hands of users.所以,我今天想谈谈 我们能够做的一些事情,也启发我们 如何改善情况、让话事权回归用户。01:52persona:n.人物角色;伪装的外表 responded:v.回答,回报;反应(respond的过去分词

10、)behavioral:adj.行为的demographic:adj.人口统计学的;人口学的 unprecedented:adj.空前的;无前例的 altruistic:adj.利他的;无私心的 goforward:发生,前进So this is Target,the company.这是塔吉特百货公司的商标。02:02I didnt just put that logo on this poor,pregnant womansbelly.我并不单单把那个商标 放在这个可怜的孕妇的肚子上。02:03You may have seen this anecdote that was printed

11、 in Forbes或许在福布斯杂志上 你看过这么一则趣事:塔吉特百货公司给这个15岁女孩寄了一份传单,传单上都TED演讲者：Jennifer Golbeck|詹尼弗.高尔拜克演讲标题：The curly fry conundrum:Why social media“likes”say more than you might think|炸薯条之迷:你以为社交媒体的“赞”就是单纯的“赞”而已吗内容概要：Do you like curly fries?Have you Liked them on Facebook?Watch this talk to find outthe surprising

12、 things Facebook(and others)can guess about you from your random Likes andShares.Computer scientist Jennifer Golbeck explains how this came about,how someapplications of the technology are not so cute-and why she thinks we should return thecontrol of information to its rightful owners.你喜欢吃炸薯条吗?你在面谱网

13、上给过它们“赞”吗?看看这个视频，你会惊奇地发现面谱网（以及其他网站）怎样从你随手的“赞”和“分享”中预测与你相关的信息。计算机科学家詹尼弗.高尔拜克解释这一种做法，解释某些科技应用软件其实并没那么趣致以及为什么她觉得我们应该把信息的控制权交还给真正的拥有人。You may have seen this anecdote that was printed in Forbesmagazine where Target sent a flyer to this 15-year-old girlwith advertisements and coupons for baby bottles andd

14、iapers and cribs two weeks before she told her parents thatshe was pregnant.百货公司给这个15岁女孩寄了一份传单,传单上都是婴儿奶瓶、尿布、婴儿床的广告和优惠券。这一切发生在 她把怀孕消息告诉父母的两周前。02:07Yeah,the dad was really upset.没错,女孩的父亲很生气。02:21He said,How did Target figure out that this high school girlwas pregnant before she told her parents?他说:”塔吉

15、特是如何 在连这个高中女生的父母都尚未知情之前 就知道她怀孕了？“02:24It turns out that they have the purchase history for hundredsof thousands of customers and they compute what they call apregnancy score,which is not just whether or not a womanspregnant,but what her due date is.原来,塔吉特有成千上万的顾客,并拥有他们的购买历史记录,他们用计算机推算出他们所谓的“怀孕分数”,不仅

16、能知道一个女性是否怀孕,而且还能计算出她的分娩日期。02:29And they compute that not by looking at the obvious things,like,shes buying a crib or baby clothes,but things like,shebought more vitamins than she normally had,or she boughta handbag thats big enough to hold diapers.他们计算出的结果 不单单是基于一些显而易见的事情,比如说,她准备买个婴儿床或孩子的衣服,更是基于其他一些

17、事情,例如她比平时多买了维他命,或她买了一个新的手提包 大得可以放尿布。02:41anecdote:n.轶事;奇闻;秘史 Forbes:n.福布斯(美国著名财经杂志);福布斯(美国出版及媒体集团);福布斯(姓氏)flyer:n.传单;飞鸟;飞行物;飞跳;孤注一掷 coupons:n.优惠券;息票;化工取样片(coupon的复数)diapers:n.尿布(diaper的复数)/vt.换尿布(diaper的第三人称单数)cribs:n.婴儿床;栅栏;食槽/vi.剽窃/vt.拘禁,关入栅栏;抄袭 pregnancy:n.怀孕;丰富,多产;意义深长crib:n.婴儿床;栅栏;食槽/vi.剽窃/vt.拘

18、禁,关入栅栏;抄袭 vitamins:n.生化维生素;生化维他命And by themselves,those purchases dont seem like theymight reveal a lot,but its a pattern of behavior that,whenyou take it in the context of thousands of other people,startsto actually reveal some insights.单独来看这些消费记录 或许并不能说明什么,但这确是一种行为模式,当你有大量人口背景作比较,这种行为模式就开始透露一些见解。0

19、2:54So thats the kind of thing that we do when were predictingstuff about you on social media.当我们根据社交媒体来预测关于你的一些事情时,这便是我们常做的一类事情。03:06Were looking for little patterns of behavior that,when youdetect them among millions of people,lets us find out allkinds of things.我们着眼于零星的行为模式,当你在众人中发现这些行为模式时,会帮助我们发

20、现各种各样的事情。03:11So in my lab and with colleagues,weve developedmechanisms where we can quite accurately predict thingslike your political preference,your personality score,gender,sexual orientation,religion,age,intelligence,along withthings like how much you trust the people you know andhow strong tho

21、se relationships are.在我的实验室,在同事们的合作下,我们已经开发了一些机制 来较为准确地推测一些事情,比如你的政治立场、你的性格得分、性别、性取向、宗教信仰、年龄、智商,另外还有:你对认识的人的信任程度、你的人际关系程度。03:19We can do all of this really well.我们能够很好地完成这些推测。03:38And again,it doesnt come from what you might think of asobvious information.我在这里在强调一遍,这种推测并基于 在你看来显而易见的信息。03:39insights:

22、n.洞察力;眼力;深刻见解(insight的复数)predicting:v.预测(predict的ing形式);预报;预言 accurately:adv.精确地,准确地 gender:n.性;性别;性交/vt.生(过去式gendered,过去分词gendered,现在分词gendering,第三人称单数genders,形容词genderless)orientation:n.方向;定向;适应;情况介绍;向东方So my favorite example is from this study that was publishedthis year in the Proceedings of the

23、 National Academies.我最喜欢的例子是来自 今年发表在 美国国家论文集上的一个研究。03:44If you Google this,youll find it.你可以在谷歌搜索找到这篇文章。03:49Its four pages,easy to read.这篇文章总共四页,容易阅读。03:50And they looked at just peoples Facebook likes,so just thethings you like on Facebook,and used that to predict allthese attributes,along with so

24、me other ones.他们仅仅研究了人们在面谱网上的“赞”,也就是你在面谱网上喜欢的事情。他们利用这些数据来预测 之前所说的所有特性,还有其他的一些特性。03:52And in their paper they listed the five likes that were mostindicative of high intelligence.在文章中列举了 最能够显示高智商的五个“赞”。04:01And among those was liking a page for curly fries.(Laughter)Curly fries are delicious,but likin

25、g them does not necessarilymean that youre smarter than the average person.在这五项中 赞“炸扭薯”页面的是其中之一 炸扭薯很好吃,但喜欢吃炸扭薯 并不一定意味着你比一般人聪明。04:07So how is it that one of the strongest indicators of yourintelligence is liking this page when the content is totallyirrelevant to the attribute thats being predicted?那么

26、为什么喜欢某个页面 就成为显示你智商 的重要因素,尽管该页面的内容和所预测的属性 与此毫不相干？04:17And it turns out that we have to look at a whole bunch ofunderlying theories to see why were able to do this.事实是我们必须审视 大量的基础理论,从而了解我们是如何做到准确推测的。04:28Proceedings:n.诉讼；行动(proceeding的复数形式)；会议记录；议程 Academies:n.院校(academy的复数形式);学院curly:adj.卷曲的;卷毛的;(木材)

27、有皱状纹理的;蜷缩的 indicators:n.指示器(indicator的复数);指示灯 irrelevant:adj.不相干的;不切题的 predicted:adj.预测的/v.预计(predict的过去分词);预示 underlying:adj.潜在的；根本的；在下面的；优先的/v.放在的下面；为的基础；优先于(underlie的ing形式)One of them is a sociological theory called homophily,whichbasically says people are friends with people like them.其中一个基础理论是社

28、会学的同质性理论,主要意思是人们和自己相似的人交朋友。04:34So if youre smart,you tend to be friends with smart people,and if youre young,you tend to be friends with youngpeople,and this is well established for hundreds of years.所以说,如果你很聪明,你倾向于和聪明的人交朋友。如果你还年轻,你倾向于和年轻人交朋友。这是数百年来 公认的理论。04:40We also know a lot about how informati

29、on spreads throughnetworks.我们很清楚 信息在网络上传播的传播途径。04:48It turns out things like viral videos or Facebook likes or otherinformation spreads in exactly the same way that diseasesspread through social networks.结果是,流行的视频、脸书上得到很多“赞”的内容、或者其他信息的传播,同疾病在社交网络中蔓延的方式是相同的。04:52So this is something weve studied for

30、a long time.我们在这方面已经研究很久了。05:01We have good models of it.我们己经建立了很好的模型。05:02And so you can put those things together and start seeingwhy things like this happen.你能够将所有这些事物放在一起,看看为什么这样的事情会发生。05:04So if I were to give you a hypothesis,it would be that a smartguy started this page,or maybe one of the fi

31、rst people wholiked it would have scored high on that test.如果要我给你一个假说的话,我会猜测一个聪明的人建立了这个页面,或者第一个喜欢这个页面的人拥有挺高的智商得分。05:09sociological:adj.社会的(等于sociologic);社会学的;针对社会问题的 spreads:n.数扩展;广阔;纺桌布(spread的复数)/v.传播(spread的第三人称单数);打开;摆放 viral:adj.滤过性毒菌引起的;滤过性毒菌的 or other:或者说 hypothesis:n.假设scored:adj.折叠的;刮伤的/v.在

32、上刻痕;牢记的宿怨;得到的分数(score的过去分词)And they liked it,and their friends saw it,and by homophily,we know that he probably had smart friends,and so it spreadto them,and some of them liked it,and they had smartfriends,and so it spread to them,and so it propagatedthrough the network to a host of smart people,so

33、that by theend,the action of liking the curly fries page is indicative ofhigh intelligence,not because of the content,but becausethe actual action of liking reflects back the commonattributes of other people who have done it.他们喜欢了这个页面,然后他们的朋友看到了,根据同质性理论,我们知道这些人可能有聪明的朋友,然后他们看到这类信息,他们中的一部分人也喜欢,他们也有聪明的

34、朋友,所以这类信息也传到其他朋友那里,所以信息就在网络上 在聪明人的圈子里流传开来了,因此到了最后,喜欢炸扭薯的这个页面 就成了高智商的象征,而不是因为内容本身,而是“喜欢”这一个实际行动 反映了那些也付诸同样行动的人 的相同特征。05:18So this is pretty complicated stuff,right?听起来很复杂,对吧？05:48Its a hard thing to sit down and explain to an average user,and even if you do,what can the average user do about it?对于一般用

35、户来说 它比较难解释清楚,就算你解释清楚了,一般用户又能利用它来干嘛呢？05:51How do you know that youve liked something that indicatesa trait for you thats totally irrelevant to the content of whatyouve liked?你又怎么能知道你喜欢的事情 反映了你什么特征而且这个特征还和你喜欢的内容毫不相干呢？05:58Theres a lot of power that users dont have to control howthis data is used.用户其实没

36、有太多的能力 去控制这些数据的使用。06:05And I see that as a real problem going forward.我把这个看作将来的真实问题,06:10propagated:传播 reflects:反照/反射 indicates:v.表明(indicate的第三人称单数形式);指示,显示 trait:n.特性,特点;品质;少许So I think theres a couple paths that we want to look at if wewant to give users some control over how this data is used,be

37、cause its not always going to be used for their benefit.我认为,要是我们想让用户拥有 使用这些数据的能力,那么有几条路径 我们需要探究,因为这些数据并不总是 用来为他们谋利益。06:13www.XiYuS锡育软件An example I often give is that,if I ever get bored being aprofessor,Im going to go start a company that predicts allof these attributes and things like how well you

38、work inteams and if youre a drug user,if youre an alcoholic.这有一个我经常举的例子,如果我厌倦了当一名教授,我会选择自己开家公司 这家公司能预测这些特性和事物 例如你在团队里的能力 例如你是否是一个吸毒者或酗酒者。06:23We know how to predict all that.我们知道如何去预测这些特性。06:33And Im going to sell reports to H.R.companies and bigbusinesses that want to hire you.然后我就会把这些报告 卖给那些人力资源公司

39、 和想要雇佣你的大公司。06:35We totally can do that now.我们完全可以做到这点。06:41I could start that business tomorrow,and you would haveabsolutely no control over me using your data like that.我明天就能开始这个项目,并且你对我这用使用你的数据 是一点办法也没有的。06:42That seems to me to be a problem.这对我来说是一个问题。06:48So one of the paths we can go down is t

40、he policy and lawpath.所以我们可选的其中一条路径 是政策和法律这条途径。06:50And in some respects,I think that that would be mosteffective,but the problem is wed actually have to do it.某程度上我觉得这可能是最有效的 但问题是,事实上我们将不得不这么做。06:54predicts:vt.预报,预言;预知/vi.作出预言;作预料,作预报 alcoholic:adj.酒精的,含酒精的/n.酒鬼,酗酒者Observing our political process i

41、n action makes me think itshighly unlikely that were going to get a bunch ofrepresentatives to sit down,learn about this,and then enact观察我们目前的政治进程 让我觉得在美国 把一帮代表们聚在一起 让他们坐下来理解这个问题,然后颁布有关知识产权法方面的颠覆性条例,让用户掌控自己的数据,07:00representatives to sit down,learn about this,and then enactsweeping changes to intell

42、ectual property law in the U.S.控自己的数据,07:00so users control their data.这似乎是不可能的。07:13We could go the policy route,where social media companiessay,you know what?You own your data.我们可以走政策途径,这样社交媒体公司就会告诉你,你知道吗？你的确拥有你的数据。07:16You have total control over how its used.你绝对能自己决定要怎么去用。07:20The problem is tha

43、t the revenue models for most socialmedia companies rely on sharing or exploiting users data insome way.但问题在于大部分的社交媒体公司 他们的盈利模式 在某方面取决于分享或挖掘用户的数据资料。07:22Its sometimes said of Facebook that the users arent thecustomer,theyre the product.所以有时会说面谱网的用户并不是顾客,07:30And so how do you get a company to cede c

44、ontrol of theirmain asset back to the users?那么你要怎样让一个公司 将他们的主要资产控制权 双手拱让给用户呢？07:34Observing:adj.观察的;注意的;观察力敏锐的/v.观察;遵守(observe的ing形式)a bunch of:一群;一束;一堆representatives:n.代表;众议院(representative的复数形式)enact:vt.颁布;制定法律;扮演;发生 sweeping:n.扫除;垃圾/adj.彻底的;广泛的;扫荡的/v.打扫;扫除(sweep的现在分词形式)rely on:依靠,依赖 exploiting:n

45、.利用;开拓/v.利用;开发(exploit的ing形式)cede:vt.放弃;割让(领土)asset:n.资产;优点;有用的东西;有利条件;财产;有价值的人或物Its possible,but I dont think its something that were goingto see change quickly.这是可能的,但我不觉得 我们能很快见证这种改变。07:41So I think the other path that we can go down thats going tobe more effective is one of more science.所以我认为我们得

46、走另一条途径 一条更有效的途径,一条更加科学的途径。07:45Its doing science that allowed us to develop all thesemechanisms for computing this personal data in the firstplace.这途径是开发一种技术 让我们能够发展所有这些机制 来首先处理自己的个人信息资料。07:50And its actually very similar research that wed have to do ifwe want to develop mechanisms that can say to a

47、 user,Heres the risk of that action you just took.而这很接近 我们必须做的研究,要是我们想要发展这些机制 跟用户说明,“这样做你需要承担那样的风险。”07:56By liking that Facebook page,or by sharing this piece ofpersonal information,youve now improved my ability topredict whether or not youre using drugs or whether or notyou get along well in the wo

48、rkplace.你在面谱网上点“赞”或者分享一些私人信息,就相当于增强了我的能力 去预测你是不是在吸毒 或者你在工作中是否顺利。08:06And that,I think,can affect whether or not people want toshare something,keep it private,or just keep it offlinealtogether.我觉得,这样做 能够影响人们分享的决定:是要保持私隐,还是在网上只字不提。08:17We can also look at things like allowing people to encryptdata tha

49、t they upload,so its kind of invisible and worthlessto sites like Facebook or third party services that access it,but that select users who the person who posted it want tosee it have access to see it.我们也可以探究一些别的,例如 让人们去给上传的东西加密,那么像面谱网这样的网站 或其他能获取信息的第三方来说 这些信息就隐秘很多,也少了很多意义,而且只有上传人指定的用户 才有浏览的权限。08:24

50、in the first place:首先;起初 get along:(勉强)生活;进展;(使)前进;与和睦相处 workplace:n.工作场所;车间 offline:n.脱机;挂线/adj.脱机的;离线的,未连线的/adv.未连线地 encrypt:vt.将译成密码 upload:vt.上传This is all super exciting research from an intellectualperspective,and so scientists are going to be willing to do it.从智能的角度来看,这是一个非常振奋人心的研究,而且科学家们也会乐意