TED@Cannes
Nicholas Christakis: How social networks predict epidemics
古樂朋：利用人際網路預測流行趨勢
Filmed:
Readability: 4.9
669,862 views
在建立了錯綜複雜的人際網路連結之後，古樂朋和同事James Fowler開始尋找運用這些資訊，促進我們生活的方法。這一次他的研究所帶來的頭條：研究人際網絡，能夠大幅提前預測流行趨勢的發生，不論是創新概念的散播，或是危險的傳染病（如H1N1）
Nicholas Christakis - Physician, social scientist
Nicholas Christakis explores how the large-scale, face-to-face social networks in which we are embedded affect our lives, and what we can do to take advantage of this fact. Full bio
Nicholas Christakis explores how the large-scale, face-to-face social networks in which we are embedded affect our lives, and what we can do to take advantage of this fact. Full bio
Double-click the English transcript below to play the video.
00:15
For the last 10 years, I've been spending my time trying to figure out
0
0
3000
過去10年來，我試著了解，
00:18
how and why human beings
1
3000
2000
人們為何形成社交網路，
00:20
assemble themselves into social networks.
2
5000
3000
以及這些網路是如何形成的。
00:23
And the kind of social network I'm talking about
3
8000
2000
我所要談的網路，
00:25
is not the recent online variety,
4
10000
2000
並非現在所謂的網路社群。
00:27
but rather, the kind of social networks
5
12000
2000
而是更原始的社交網路，
00:29
that human beings have been assembling for hundreds of thousands of years,
6
14000
3000
自從人類在非洲大草原出現以來，
00:32
ever since we emerged from the African savannah.
7
17000
3000
已經使用這種連結十幾萬年了。
00:35
So, I form friendships and co-worker
8
20000
2000
我和其他人分享友誼、同事、
00:37
and sibling and relative relationships with other people
9
22000
3000
手足和親戚等等人際關係，
00:40
who in turn have similar relationships with other people.
10
25000
2000
這些人也和其他人有相似的連結。
00:42
And this spreads on out endlessly into a distance.
11
27000
3000
這樣的連結向外擴散，
00:45
And you get a network that looks like this.
12
30000
2000
從而得到的網路看起來會像這樣。
00:47
Every dot is a person.
13
32000
2000
每點代表一個人，
00:49
Every line between them is a relationship between two people --
14
34000
2000
兩點間的線則代表兩個人之間的關係，
00:51
different kinds of relationships.
15
36000
2000
各種不同的關係。
00:53
And you can get this kind of vast fabric of humanity,
16
38000
3000
種種的關係交織成一幅巨大的網路，
00:56
in which we're all embedded.
17
41000
2000
而我們都位於其中。
00:58
And my colleague, James Fowler and I have been studying for quite sometime
18
43000
3000
我的同事James Fowler和我花了滿長時間研究，
01:01
what are the mathematical, social,
19
46000
2000
想找到一個基於數學、社會學、
01:03
biological and psychological rules
20
48000
3000
生物學或是心理學的規則，
01:06
that govern how these networks are assembled
21
51000
2000
能夠主導這些網路的形成。
01:08
and what are the similar rules
22
53000
2000
以及是否有類似的規則
01:10
that govern how they operate, how they affect our lives.
23
55000
3000
主導網路的運作，進而影響我們的生活。
01:13
But recently, we've been wondering
24
58000
2000
直到最近，我們開始思考，
01:15
whether it might be possible to take advantage of this insight,
25
60000
3000
是否有可能利用這些發現，
01:18
to actually find ways to improve the world,
26
63000
2000
來找出增進人類福祉的方法，
01:20
to do something better,
27
65000
2000
改善現況，
01:22
to actually fix things, not just understand things.
28
67000
3000
去導正，而非只是單純理解問題。
01:25
So one of the first things we thought we would tackle
29
70000
3000
我們最先著手研究的議題，
01:28
would be how we go about predicting epidemics.
30
73000
3000
是如何預測流行趨勢。
01:31
And the current state of the art in predicting an epidemic --
31
76000
2000
目前最先進的預測方法—
01:33
if you're the CDC or some other national body --
32
78000
3000
如果你在疾病管制中心(CDC)或類似的政府單位工作—
01:36
is to sit in the middle where you are
33
81000
2000
是待在中央枯等，
01:38
and collect data
34
83000
2000
並收集資料，
01:40
from physicians and laboratories in the field
35
85000
2000
第一線的醫生和實驗室把資料傳進來，
01:42
that report the prevalence or the incidence of certain conditions.
36
87000
3000
報告疾病的流行程度或發生機率。
01:45
So, so and so patients have been diagnosed with something,
37
90000
3000
這邊有某個病患被診斷出來，
01:48
or other patients have been diagnosed,
38
93000
2000
那邊又有別人得病。
01:50
and all these data are fed into a central repository, with some delay.
39
95000
3000
資訊經過一些延遲之後，傳進中央的資料庫裡。
01:53
And if everything goes smoothly,
40
98000
2000
如果一切順利，
01:55
one to two weeks from now
41
100000
2000
一到兩個禮拜之後，
01:57
you'll know where the epidemic was today.
42
102000
3000
我們才會得知當天流行病的狀況。
02:00
And actually, about a year or so ago,
43
105000
2000
事實上一年多以前，
02:02
there was this promulgation
44
107000
2000
有人發表了這樣的概念，
02:04
of the idea of Google Flu Trends, with respect to the flu,
45
109000
3000
使用Google流感趨勢(Flu Trends)來尋找流感。
02:07
where by looking at people's searching behavior today,
46
112000
3000
透過對搜尋行為的分析，
02:10
we could know where the flu --
47
115000
2000
我們能夠得知流感發生的區域，
02:12
what the status of the epidemic was today,
48
117000
2000
得知當天傳染病的狀態，
02:14
what's the prevalence of the epidemic today.
49
119000
3000
以及傳染病的影響程度。
02:17
But what I'd like to show you today
50
122000
2000
不過這次我要介紹的方法，
02:19
is a means by which we might get
51
124000
2000
讓我們不只能夠
02:21
not just rapid warning about an epidemic,
52
126000
3000
得到傳染病的快速預警，
02:24
but also actually
53
129000
2000
更能夠讓我們
02:26
early detection of an epidemic.
54
131000
2000
提早偵測到流行病的發生。
02:28
And, in fact, this idea can be used
55
133000
2000
事實上，這個概念不止能夠
02:30
not just to predict epidemics of germs,
56
135000
3000
用來預測病菌的流行，
02:33
but also to predict epidemics of all sorts of kinds.
57
138000
3000
也能夠應用來預測各種事物的趨勢。
02:37
For example, anything that spreads by a form of social contagion
58
142000
3000
例如，任何能透過社群的方式傳播的事物，
02:40
could be understood in this way,
59
145000
2000
都可以用這種方式理解。
02:42
from abstract ideas on the left
60
147000
2000
從左邊的抽象概念，
02:44
like patriotism, or altruism, or religion
61
149000
3000
像是愛國主義、利他精神，或是宗教，
02:47
to practices
62
152000
2000
到具體的事物，
02:49
like dieting behavior, or book purchasing,
63
154000
2000
像是飲食行為、購買書籍、
02:51
or drinking, or bicycle-helmet [and] other safety practices,
64
156000
3000
酗酒、使用腳踏車安全帽等安全措施，
02:54
or products that people might buy,
65
159000
2000
或是日常用品，
02:56
purchases of electronic goods,
66
161000
2000
電子產品，
02:58
anything in which there's kind of an interpersonal spread.
67
163000
3000
任何透過人與人之間傳遞的事物。
03:01
A kind of a diffusion of innovation
68
166000
2000
這種創新的擴散，
03:03
could be understood and predicted
69
168000
2000
可以透過接下來我將展示的機制，
03:05
by the mechanism I'm going to show you now.
70
170000
3000
來理解並且預測。
03:08
So, as all of you probably know,
71
173000
2000
你們或許知道，
03:10
the classic way of thinking about this
72
175000
2000
最經典的範例，
03:12
is the diffusion-of-innovation,
73
177000
2000
就是創新的擴散，
03:14
or the adoption curve.
74
179000
2000
或是所謂的「普及曲線」。
03:16
So here on the Y-axis, we have the percent of the people affected,
75
181000
2000
Y軸是受影響人數的百分比，
03:18
and on the X-axis, we have time.
76
183000
2000
X軸表示時間的推移。
03:20
And at the very beginning, not too many people are affected,
77
185000
3000
剛開始沒有太多人受到影響，
03:23
and you get this classic sigmoidal,
78
188000
2000
然後你會看到經典的反曲線，
03:25
or S-shaped, curve.
79
190000
2000
或是S型曲線。
03:27
And the reason for this shape is that at the very beginning,
80
192000
2000
形成這種曲線的原因是，
03:29
let's say one or two people
81
194000
2000
一開始只有一兩個人
03:31
are infected, or affected by the thing
82
196000
2000
被影響，或是被「感染」，
03:33
and then they affect, or infect, two people,
83
198000
2000
然後傳遞給另外兩個人，
03:35
who in turn affect four, eight, 16 and so forth,
84
200000
3000
接著4、8、16，以此類推，
03:38
and you get the epidemic growth phase of the curve.
85
203000
3000
這時進入迅速增長的階段。
03:41
And eventually, you saturate the population.
86
206000
2000
最終擴散到整個群體。
03:43
There are fewer and fewer people
87
208000
2000
於是越來越難找到
03:45
who are still available that you might infect,
88
210000
2000
尚未被影響的人，
03:47
and then you get the plateau of the curve,
89
212000
2000
這時候曲線進入高原期，
03:49
and you get this classic sigmoidal curve.
90
214000
3000
形成整條反曲線。
03:52
And this holds for germs, ideas,
91
217000
2000
這個模式在病菌、創意、
03:54
product adoption, behaviors,
92
219000
2000
新產品的普及、行為，
03:56
and the like.
93
221000
2000
以及類似情況都適用。
03:58
But things don't just diffuse in human populations at random.
94
223000
3000
要注意的是，事物並不是隨機在人群中蔓延，
04:01
They actually diffuse through networks.
95
226000
2000
而是隨著網路分布來擴散。
04:03
Because, as I said, we live our lives in networks,
96
228000
3000
因為我們活在網路的世界，
04:06
and these networks have a particular kind of a structure.
97
231000
3000
而這種網路有特定的結構。
04:09
Now if you look at a network like this --
98
234000
2000
觀察這個網路，
04:11
this is 105 people.
99
236000
2000
裡面有105人。
04:13
And the lines represent -- the dots are the people,
100
238000
2000
每個點代表一個人
04:15
and the lines represent friendship relationships.
101
240000
2000
每條線代表彼此間的友誼關係。
04:17
You might see that people occupy
102
242000
2000
人們在這個網路中
04:19
different locations within the network.
103
244000
2000
佔據不同的位置，
04:21
And there are different kinds of relationships between the people.
104
246000
2000
彼此間有不同類型的關係。
04:23
You could have friendship relationships, sibling relationships,
105
248000
3000
可能是朋友、手足、
04:26
spousal relationships, co-worker relationships,
106
251000
3000
配偶、同事、
04:29
neighbor relationships and the like.
107
254000
3000
鄰居等等。
04:32
And different sorts of things
108
257000
2000
不同的事物會
04:34
spread across different sorts of ties.
109
259000
2000
透過不同的關係來傳播。
04:36
For instance, sexually transmitted diseases
110
261000
2000
例如，性傳染病，
04:38
will spread across sexual ties.
111
263000
2000
會藉由性伴侶的聯繫來散佈。
04:40
Or, for instance, people's smoking behavior
112
265000
2000
或者像人們吸菸，
04:42
might be influenced by their friends.
113
267000
2000
可能是受到朋友的影響。
04:44
Or their altruistic or their charitable giving behavior
114
269000
2000
人們的善行或捐助，
04:46
might be influenced by their coworkers,
115
271000
2000
可能是出自同事間的影響，
04:48
or by their neighbors.
116
273000
2000
或是他們鄰居的行為。
04:50
But not all positions in the network are the same.
117
275000
3000
但是網路中的位置並非都一樣。
04:53
So if you look at this, you might immediately grasp
118
278000
2000
這張圖或許能讓你了解，
04:55
that different people have different numbers of connections.
119
280000
3000
不同人有不同數量的連結。
04:58
Some people have one connection, some have two,
120
283000
2000
有的人一個，有人兩個，
05:00
some have six, some have 10 connections.
121
285000
3000
有人六個，有的人擁有十個連結。
05:03
And this is called the "degree" of a node,
122
288000
2000
也就是一個節點的「度數」，
05:05
or the number of connections that a node has.
123
290000
2000
或是一個節點所擁有的連結數。
05:07
But in addition, there's something else.
124
292000
2000
除此之外，
05:09
So, if you look at nodes A and B,
125
294000
2000
如果觀察節點A與B，
05:11
they both have six connections.
126
296000
2000
兩者都擁有六個連結。
05:13
But if you can see this image [of the network] from a bird's eye view,
127
298000
3000
但是如果鳥瞰整個圖像，
05:16
you can appreciate that there's something very different
128
301000
2000
你就會發現兩者之間，
05:18
about nodes A and B.
129
303000
2000
A與B的不同之處
05:20
So, let me ask you this -- I can cultivate this intuition by asking a question --
130
305000
3000
問題來了 －請用直覺回答－
05:23
who would you rather be
131
308000
2000
你比較想當誰：
05:25
if a deadly germ was spreading through the network, A or B?
132
310000
3000
如果致命病菌正在網路中散佈，A或是B？
05:28
(Audience: B.) Nicholas Christakis: B, it's obvious.
133
313000
2000
（觀眾：B）很明顯的是B。
05:30
B is located on the edge of the network.
134
315000
2000
B處在網路的邊緣。
05:32
Now, who would you rather be
135
317000
2000
現在，你比較想當誰：
05:34
if a juicy piece of gossip were spreading through the network?
136
319000
3000
如果網路中流傳著一個天大的八卦？
05:37
A. And you have an immediate appreciation
137
322000
3000
A。而且你馬上能夠理解到，
05:40
that A is going to be more likely
138
325000
2000
A會有更高的機率
05:42
to get the thing that's spreading and to get it sooner
139
327000
3000
趕上流行，而且早先一步。
05:45
by virtue of their structural location within the network.
140
330000
3000
這要歸功於他們在網路中的位置。
05:48
A, in fact, is more central,
141
333000
2000
A比較靠近中央，
05:50
and this can be formalized mathematically.
142
335000
3000
這可以用數學形式來描述。
05:53
So, if we want to track something
143
338000
2000
因此，如果我們希望追蹤某些事物
05:55
that was spreading through a network,
144
340000
3000
在網路中散佈的狀態，
05:58
what we ideally would like to do is to set up sensors
145
343000
2000
理想狀況是佈置感測器，
06:00
on the central individuals within the network,
146
345000
2000
對準網路裡的中央個體，
06:02
including node A,
147
347000
2000
包括節點A。
06:04
monitor those people that are right there in the middle of the network,
148
349000
3000
監視這些位於中心位置的人們，
06:07
and somehow get an early detection
149
352000
2000
以早期的預警到
06:09
of whatever it is that is spreading through the network.
150
354000
3000
正在網路上傳播的事物。
06:12
So if you saw them contract a germ or a piece of information,
151
357000
3000
亦即，如果這些人染病或是獲悉某些資訊，
06:15
you would know that, soon enough,
152
360000
2000
你就可以推斷，要不了多久，
06:17
everybody was about to contract this germ
153
362000
2000
所有人都會被波及，不管是染病，
06:19
or this piece of information.
154
364000
2000
或是得到資訊。
06:21
And this would be much better
155
366000
2000
這樣的作法遠勝於
06:23
than monitoring six randomly chosen people,
156
368000
2000
隨機挑選六個人來監控，
06:25
without reference to the structure of the population.
157
370000
3000
因為該做法並未考慮到群體的結構。
06:28
And in fact, if you could do that,
158
373000
2000
若是真的能夠實行，
06:30
what you would see is something like this.
159
375000
2000
我們會得到類似這樣的情況：
06:32
On the left-hand panel, again, we have the S-shaped curve of adoption.
160
377000
3000
左邊的圖表，是S型的普及曲線。
06:35
In the dotted red line, we show
161
380000
2000
我們用紅色虛線標示出，
06:37
what the adoption would be in the random people,
162
382000
2000
一般人的普及情形，
06:39
and in the left-hand line, shifted to the left,
163
384000
3000
左邊的線段，則向左偏移，
06:42
we show what the adoption would be
164
387000
2000
顯示出網路中的核心個體，
06:44
in the central individuals within the network.
165
389000
2000
他們的普及情形。
06:46
On the Y-axis is the cumulative instances of contagion,
166
391000
2000
Y軸是受到傳染「病例」的累積數量，
06:48
and on the X-axis is the time.
167
393000
2000
X軸則是時間。
06:50
And on the right-hand side, we show the same data,
168
395000
2000
右邊的圖表是相同的資料，
06:52
but here with daily incidence.
169
397000
2000
呈現的是每日的「感染」數字。
06:54
And what we show here is -- like, here --
170
399000
2000
我們想要傳達的是，
06:56
very few people are affected, more and more and more and up to here,
171
401000
2000
一開始少數人受到影響，然後越來越多直到這裡，
06:58
and here's the peak of the epidemic.
172
403000
2000
這裡就是傳播的高峰期。
07:00
But shifted to the left is what's occurring in the central individuals.
173
405000
2000
向左偏的則是在核心個體發生的情形，
07:02
And this difference in time between the two
174
407000
3000
這兩條曲線間的時間差，
07:05
is the early detection, the early warning we can get,
175
410000
3000
就是預測時差，我們可以從中得到預警，
07:08
about an impending epidemic
176
413000
2000
人群中是否有
07:10
in the human population.
177
415000
2000
即將爆發的疫情。
07:12
The problem, however,
178
417000
2000
然而問題在於，
07:14
is that mapping human social networks
179
419000
2000
人際間的社交網路，
07:16
is not always possible.
180
421000
2000
並不容易繪測。
07:18
It can be expensive, not feasible,
181
423000
2000
這樣的計畫可能所費不貲、非常困難、
07:20
unethical,
182
425000
2000
具有道德爭議
07:22
or, frankly, just not possible to do such a thing.
183
427000
3000
說實話，就是不可能。
07:25
So, how can we figure out
184
430000
2000
所以，我們要如何找出，
07:27
who the central people are in a network
185
432000
2000
網路中的核心個體在哪，
07:29
without actually mapping the network?
186
434000
3000
而無需繪出整個網路？
07:32
What we came up with
187
437000
2000
我們所想到的，
07:34
was an idea to exploit an old fact,
188
439000
2000
是利用一個既有的事實
07:36
or a known fact, about social networks,
189
441000
2000
關於社交網路，眾所皆知的事實。
07:38
which goes like this:
190
443000
2000
也就是：
07:40
Do you know that your friends
191
445000
2000
你知道你的朋友，
07:42
have more friends than you do?
192
447000
3000
所擁有的友人數目比你還多嗎？
07:45
Your friends have more friends than you do,
193
450000
3000
朋友的友人數目比自己擁有的還多，
07:48
and this is known as the friendship paradox.
194
453000
2000
通常這種情況被稱做「友誼悖論」。
07:50
Imagine a very popular person in the social network --
195
455000
2000
試想社交網路中的人氣王 －
07:52
like a party host who has hundreds of friends --
196
457000
3000
例如派對的主人，身邊有上百個朋友 --
07:55
and a misanthrope who has just one friend,
197
460000
2000
和孤僻成性，只有一個朋友的人。
07:57
and you pick someone at random from the population;
198
462000
3000
若是你隨便從人群中挑出一位，
08:00
they were much more likely to know the party host.
199
465000
2000
他們就非常有可能認識這位派對主人，
08:02
And if they nominate the party host as their friend,
200
467000
2000
而當他們舉出派對主人是自己的朋友，
08:04
that party host has a hundred friends,
201
469000
2000
由於他有上百個朋友，
08:06
therefore, has more friends than they do.
202
471000
3000
因此遠比自己的朋友數目還多。
08:09
And this, in essence, is what's known as the friendship paradox.
203
474000
3000
在本質上，這就是友誼悖論：
08:12
The friends of randomly chosen people
204
477000
3000
隨機挑選的人，他的朋友，
08:15
have higher degree, and are more central
205
480000
2000
會有較高的連結數目，也較為趨近核心，
08:17
than the random people themselves.
206
482000
2000
因而優於那些隨機挑選的人。
08:19
And you can get an intuitive appreciation for this
207
484000
2000
因此，你可以憑直覺想像，
08:21
if you imagine just the people at the perimeter of the network.
208
486000
3000
如果是那些位於網路邊緣的人，
08:24
If you pick this person,
209
489000
2000
這樣的人，
08:26
the only friend they have to nominate is this person,
210
491000
3000
他的朋友只會有這個人，
08:29
who, by construction, must have at least two
211
494000
2000
而結構上來說，這個人至少會有兩位、
08:31
and typically more friends.
212
496000
2000
甚至更多的朋友。
08:33
And that happens at every peripheral node.
213
498000
2000
在每個外圍的節點都是這樣。
08:35
And in fact, it happens throughout the network as you move in,
214
500000
3000
當你越往網路的中心移動時就越常見，
08:38
everyone you pick, when they nominate a random --
215
503000
2000
每個被你挑到的人，當他們隨意提出一個...
08:40
when a random person nominates a friend of theirs,
216
505000
3000
每當提出一個他們的朋友，
08:43
you move closer to the center of the network.
217
508000
3000
你就越靠近網路的中心。
08:46
So, we thought we would exploit this idea
218
511000
3000
於是我們認為可以利用這個概念，
08:49
in order to study whether we could predict phenomena within networks.
219
514000
3000
來研究我們是否能預測網路中所發生的現象。
08:52
Because now, with this idea
220
517000
2000
因為有了這樣的發現，
08:54
we can take a random sample of people,
221
519000
2000
我們可以從人群中隨機挑選樣本，
08:56
have them nominate their friends,
222
521000
2000
請他們指出他們的朋友，
08:58
those friends would be more central,
223
523000
2000
這些朋友會比較靠近中心，
09:00
and we could do this without having to map the network.
224
525000
3000
而我們就無須標出整個網路的圖像。
09:03
And we tested this idea with an outbreak of H1N1 flu
225
528000
3000
在哈佛大學，我們利用H1N1流感的爆發
09:06
at Harvard College
226
531000
2000
來測試這個概念。
09:08
in the fall and winter of 2009, just a few months ago.
227
533000
3000
在2009年秋冬，只有幾個月前，
09:11
We took 1,300 randomly selected undergraduates,
228
536000
3000
我們隨機挑選了1300位大學生，
09:14
we had them nominate their friends,
229
539000
2000
請這些人提供他們的朋友名單，
09:16
and we followed both the random students and their friends
230
541000
2000
我們同時追蹤了這些人和他們的朋友，
09:18
daily in time
231
543000
2000
每天為間隔，
09:20
to see whether or not they had the flu epidemic.
232
545000
3000
確認他們是否染上流感。
09:23
And we did this passively by looking at whether or not they'd gone to university health services.
233
548000
3000
除了被動觀察他們是否去健康中心報到，
09:26
And also, we had them [actively] email us a couple of times a week.
234
551000
3000
同時也要求每個禮拜Email給我們。
09:29
Exactly what we predicted happened.
235
554000
3000
結果一如我們所預期。
09:32
So the random group is in the red line.
236
557000
3000
隨機挑選的群體用紅線標示，
09:35
The epidemic in the friends group has shifted to the left, over here.
237
560000
3000
他們的朋友則向左邊偏移，在這邊。
09:38
And the difference in the two is 16 days.
238
563000
3000
兩者間的差距是16天。
09:41
By monitoring the friends group,
239
566000
2000
觀察朋友的群體，
09:43
we could get 16 days advance warning
240
568000
2000
能夠讓我們提早16天得到警示，
09:45
of an impending epidemic in this human population.
241
570000
3000
警告人群中即將爆發的傳染病。
09:48
Now, in addition to that,
242
573000
2000
除此之外，
09:50
if you were an analyst who was trying to study an epidemic
243
575000
3000
如果你是研究傳染病的分析師，
09:53
or to predict the adoption of a product, for example,
244
578000
3000
或者想要預測產品的普及情形。
09:56
what you could do is you could pick a random sample of the population,
245
581000
3000
你可以從人群中挑選隨機樣本，
09:59
also have them nominate their friends and follow the friends
246
584000
3000
請他們指出自己的朋友，
10:02
and follow both the randoms and the friends.
247
587000
3000
並且同時追蹤這兩群樣本("隨機群"和"朋友群")。
10:05
Among the friends, the first evidence you saw of a blip above zero
248
590000
3000
在朋友群中，當曲線首次開始上升...
10:08
in adoption of the innovation, for example,
249
593000
3000
...例如創新概念的普及，
10:11
would be evidence of an impending epidemic.
250
596000
2000
這個轉折便能標示出即將發生的流行趨勢。
10:13
Or you could see the first time the two curves diverged,
251
598000
3000
另一種情況是當兩條曲線首次出現分歧時，
10:16
as shown on the left.
252
601000
2000
如左圖所示。
10:18
When did the randoms -- when did the friends take off
253
603000
3000
隨機群...他們的朋友群是何時起頭，
10:21
and leave the randoms,
254
606000
2000
離開隨機群的曲線，
10:23
and [when did] their curve start shifting?
255
608000
2000
使得這條曲線開始偏移？
10:25
And that, as indicated by the white line,
256
610000
2000
從白線上可以發現，
10:27
occurred 46 days
257
612000
2000
在整體趨勢達到高峰之前，
10:29
before the peak of the epidemic.
258
614000
2000
提早了46天。
10:31
So this would be a technique
259
616000
2000
這樣的技術，
10:33
whereby we could get more than a month-and-a-half warning
260
618000
2000
可以讓我們提早一個半月得到預警，
10:35
about a flu epidemic in a particular population.
261
620000
3000
得知特定群體中感冒的流行。
10:38
I should say that
262
623000
2000
應該這樣說，
10:40
how far advanced a notice one might get about something
263
625000
2000
我們能夠多早預知事件的發生，
10:42
depends on a host of factors.
264
627000
2000
取決於幾個主要的因素。
10:44
It could depend on the nature of the pathogen --
265
629000
2000
可能由於病原的性質 －
10:46
different pathogens,
266
631000
2000
不同的病原體，
10:48
using this technique, you'd get different warning --
267
633000
2000
利用這個技術，可以得到不同的警示 －
10:50
or other phenomena that are spreading,
268
635000
2000
或是可以說，在人際網路的結構裡
10:52
or frankly, on the structure of the human network.
269
637000
3000
某些正在傳播中的現象。
10:55
Now in our case, although it wasn't necessary,
270
640000
3000
雖然並非必要，不過在這個案例中，
10:58
we could also actually map the network of the students.
271
643000
2000
我們可以將學生的網路完整描繪出來，
11:00
So, this is a map of 714 students
272
645000
2000
所以，這幅圖包含了714個學生，
11:02
and their friendship ties.
273
647000
2000
以及他們的人際關係。
11:04
And in a minute now, I'm going to put this map into motion.
274
649000
2000
接下來我會用動畫呈現這幅圖，
11:06
We're going to take daily cuts through the network
275
651000
2000
逐日推進，
11:08
for 120 days.
276
653000
2000
一共120天。
11:10
The red dots are going to be cases of the flu,
277
655000
3000
紅點代表受到感染的案例，
11:13
and the yellow dots are going to be friends of the people with the flu.
278
658000
3000
黃點則代表受感染學生的朋友，
11:16
And the size of the dots is going to be proportional
279
661000
2000
而點的大小則以比例的方式，
11:18
to how many of their friends have the flu.
280
663000
2000
呈現它周遭朋友受到傳染的數量，
11:20
So bigger dots mean more of your friends have the flu.
281
665000
3000
也就是說，越大的點代表你有越多的朋友感冒。
11:23
And if you look at this image -- here we are now in September the 13th --
282
668000
3000
觀察這張圖 －現在是9月13號－
11:26
you're going to see a few cases light up.
283
671000
2000
你會看到幾個病例亮起來。
11:28
You're going to see kind of blooming of the flu in the middle.
284
673000
2000
中心區域裡，傳染就像開花一樣向外散布。
11:30
Here we are on October the 19th.
285
675000
3000
接下來到了10月19號，
11:33
The slope of the epidemic curve is approaching now, in November.
286
678000
2000
傳染曲線開始上升，到了11月，
11:35
Bang, bang, bang, bang, bang -- you're going to see lots of blooming in the middle,
287
680000
3000
砰，砰，砰，越來越多病例在中央區域發生。
11:38
and then you're going to see a sort of leveling off,
288
683000
2000
接著情勢開始趨緩，
11:40
fewer and fewer cases towards the end of December.
289
685000
3000
越來越少人受到感染，直到十二月底。
11:43
And this type of a visualization
290
688000
2000
這種類型的圖像化資訊，
11:45
can show that epidemics like this take root
291
690000
2000
可以呈現出流行事件開始扎根，
11:47
and affect central individuals first,
292
692000
2000
先影響中心的個體，
11:49
before they affect others.
293
694000
2000
再向外擴散的全貌。
11:51
Now, as I've been suggesting,
294
696000
2000
如我之前所說的，
11:53
this method is not restricted to germs,
295
698000
3000
這套方法並不局限於病菌，
11:56
but actually to anything that spreads in populations.
296
701000
2000
可以是透過人群傳播的任何事物。
11:58
Information spreads in populations,
297
703000
2000
資訊透過人群傳遞，
12:00
norms can spread in populations,
298
705000
2000
規則能透過人群來散佈，
12:02
behaviors can spread in populations.
299
707000
2000
行為也能夠透過人群傳播
12:04
And by behaviors, I can mean things like criminal behavior,
300
709000
3000
談到行為，像是犯罪，
12:07
or voting behavior, or health care behavior,
301
712000
3000
投票，衛生習慣－
12:10
like smoking, or vaccination,
302
715000
2000
像是吸菸或是疫苗接種，
12:12
or product adoption, or other kinds of behaviors
303
717000
2000
新產品的採用，或是其他種類的行為，
12:14
that relate to interpersonal influence.
304
719000
2000
與人們之間的相互影響有關。
12:16
If I'm likely to do something that affects others around me,
305
721000
3000
如果我打算做某些事來影響周圍的人，
12:19
this technique can get early warning or early detection
306
724000
3000
這套技巧就可以提前預警，或是偵測，
12:22
about the adoption within the population.
307
727000
3000
事物在人群中的普及程度。
12:25
The key thing is that for it to work,
308
730000
2000
讓它管用的關鍵在於，
12:27
there has to be interpersonal influence.
309
732000
2000
人們之間要能互相影響，
12:29
It cannot be because of some broadcast mechanism
310
734000
2000
而非因為某種廣播機制，
12:31
affecting everyone uniformly.
311
736000
3000
使得每個人都受到相同的影響。
12:35
Now the same insights
312
740000
2000
同樣的發現，
12:37
can also be exploited -- with respect to networks --
313
742000
3000
透過網路的傳播，也能夠有
12:40
can also be exploited in other ways,
314
745000
3000
各式各樣的應用，
12:43
for example, in the use of targeting
315
748000
2000
例如，用來標示出，
12:45
specific people for interventions.
316
750000
2000
特定的目標以進行干預。
12:47
So, for example, most of you are probably familiar
317
752000
2000
舉例來說，大部分的人可能對
12:49
with the notion of herd immunity.
318
754000
2000
"群體免疫力"感到熟悉。
12:51
So, if we have a population of a thousand people,
319
756000
3000
如果這裡有一千人的群體，
12:54
and we want to make the population immune to a pathogen,
320
759000
3000
我們希望讓群體對某個病原體免疫，
12:57
we don't have to immunize every single person.
321
762000
2000
我們不需要對每個人施打疫苗。
12:59
If we immunize 960 of them,
322
764000
2000
若是讓其中960人免疫，
13:01
it's as if we had immunized a hundred [percent] of them.
323
766000
3000
效果就相當於整個群體都免疫，
13:04
Because even if one or two of the non-immune people gets infected,
324
769000
3000
因為即使一兩個沒有免疫能力的人受到感染，
13:07
there's no one for them to infect.
325
772000
2000
他也沒有人能夠傳染，
13:09
They are surrounded by immunized people.
326
774000
2000
感染者被免疫的人所圍繞。
13:11
So 96 percent is as good as 100 percent.
327
776000
3000
所以百分之96的效果相當於百分之百。
13:14
Well, some other scientists have estimated
328
779000
2000
其他的科學家估計，
13:16
what would happen if you took a 30 percent random sample
329
781000
2000
如果只靠30%的隨機樣本，
13:18
of these 1000 people, 300 people and immunized them.
330
783000
3000
30%在1000人中，也就是讓300個人免疫，
13:21
Would you get any population-level immunity?
331
786000
2000
是否能夠達到群體層次的免疫？
13:23
And the answer is no.
332
788000
3000
答案是"不能"。
13:26
But if you took this 30 percent, these 300 people
333
791000
2000
但是，如果對這30%，要300個人
13:28
and had them nominate their friends
334
793000
2000
舉出他們的朋友，
13:30
and took the same number of vaccine doses
335
795000
3000
然後用同樣數量的疫苗藥劑，
13:33
and vaccinated the friends of the 300 --
336
798000
2000
為這群300人的朋友接種，
13:35
the 300 friends --
337
800000
2000
300位朋友，
13:37
you can get the same level of herd immunity
338
802000
2000
就能夠得到相同於，讓96%的人免疫
13:39
as if you had vaccinated 96 percent of the population
339
804000
3000
所達到的群體免疫程度。
13:42
at a much greater efficiency, with a strict budget constraint.
340
807000
3000
更有效率，也節省預算。
13:45
And similar ideas can be used, for instance,
341
810000
2000
類似的概念也能用於
13:47
to target distribution of things like bed nets
342
812000
2000
物資的分配標的，例如在發展中國家
13:49
in the developing world.
343
814000
2000
蚊帳的分發方式。
13:51
If we could understand the structure of networks in villages,
344
816000
3000
若是能夠了解村落中的網路架構，
13:54
we could target to whom to give the interventions
345
819000
2000
我們就能影響關鍵的節點，
13:56
to foster these kinds of spreads.
346
821000
2000
以增進這種形式的散佈。
13:58
Or, frankly, for advertising with all kinds of products.
347
823000
3000
或是老實說，用來宣傳各式各樣的產品。
14:01
If we could understand how to target,
348
826000
2000
如果能夠了解
14:03
it could affect the efficiency
349
828000
2000
如何鎖定焦點，
14:05
of what we're trying to achieve.
350
830000
2000
就可以提高成功的效率。
14:07
And in fact, we can use data
351
832000
2000
事實上現在有數不清的來源。
14:09
from all kinds of sources nowadays [to do this].
352
834000
2000
能夠提供我們所需的資料。
14:11
This is a map of eight million phone users
353
836000
2000
這是一份歐洲國家中，
14:13
in a European country.
354
838000
2000
八百萬電話用戶的分布圖。
14:15
Every dot is a person, and every line represents
355
840000
2000
每個點代表一個用戶，每條線
14:17
a volume of calls between the people.
356
842000
2000
代表人們之間的通話量。
14:19
And we can use such data, that's being passively obtained,
357
844000
3000
我們可以利用這份被動獲得的資料，
14:22
to map these whole countries
358
847000
2000
描繪出整個國家的全貌，
14:24
and understand who is located where within the network.
359
849000
3000
並且定位每個人在網路中的位置，
14:27
Without actually having to query them at all,
360
852000
2000
而無須一個個去問，
14:29
we can get this kind of a structural insight.
361
854000
2000
從而得到對整體架構的瞭解。
14:31
And other sources of information, as you're no doubt aware
362
856000
3000
你一定也知道，其他來源的資訊
14:34
are available about such features, from email interactions,
363
859000
3000
也能提供類似的特徵，從email互動，
14:37
online interactions,
364
862000
2000
線上互動，
14:39
online social networks and so forth.
365
864000
3000
線上社群網站等等。
14:42
And in fact, we are in the era of what I would call
366
867000
2000
而事實上我們正在這樣的一個世界，
14:44
"massive-passive" data collection efforts.
367
869000
3000
「巨量-被動」的資料被收集起來。
14:47
They're all kinds of ways we can use massively collected data
368
872000
3000
我們有一大堆方法可以使用這些廣泛收集的資料，
14:50
to create sensor networks
369
875000
3000
用來建立偵測網路，
14:53
to follow the population,
370
878000
2000
用來追蹤人群，
14:55
understand what's happening in the population,
371
880000
2000
找出群體中正在發生的事件，
14:57
and intervene in the population for the better.
372
882000
3000
並且適時介入以改善情況。
15:00
Because these new technologies tell us
373
885000
2000
因為這些新的科技讓我們理解，
15:02
not just who is talking to whom,
374
887000
2000
不只是誰正和誰溝通，
15:04
but where everyone is,
375
889000
2000
還有每個人的位置所在。
15:06
and what they're thinking based on what they're uploading on the Internet,
376
891000
3000
人們在想什麼，是看他們上傳了什麼到網路上，
15:09
and what they're buying based on their purchases.
377
894000
2000
現在的購買決策受到過去購物的影響。
15:11
And all this administrative data can be pulled together
378
896000
3000
所有這樣的資料可以組織起來，
15:14
and processed to understand human behavior
379
899000
2000
經過處理以了解人類的行為，
15:16
in a way we never could before.
380
901000
3000
以一種前所未見的方式。
15:19
So, for example, we could use truckers' purchases of fuel.
381
904000
3000
舉例來說，我們可以觀察卡車司機加油，
15:22
So the truckers are just going about their business,
382
907000
2000
司機們正準備開工，
15:24
and they're buying fuel.
383
909000
2000
他們買入了汽油，
15:26
And we see a blip up in the truckers' purchases of fuel,
384
911000
3000
我們觀察到卡車司機加油的曲線開始上升，
15:29
and we know that a recession is about to end.
385
914000
2000
而能夠推估景氣即將好轉了。
15:31
Or we can monitor the velocity
386
916000
2000
或是可以透過手機，
15:33
with which people are moving with their phones on a highway,
387
918000
3000
監視高速公路上人們的移動速度，
15:36
and the phone company can see,
388
921000
2000
電信公司便能夠得知，
15:38
as the velocity is slowing down,
389
923000
2000
當移動速度下降的時候，
15:40
that there's a traffic jam.
390
925000
2000
代表可能有交通堵塞。
15:42
And they can feed that information back to their subscribers,
391
927000
3000
這些資訊便回傳給電信公司的用戶，
15:45
but only to their subscribers on the same highway
392
930000
2000
並且針對那些在同一條高速公路上，
15:47
located behind the traffic jam!
393
932000
2000
位於車陣後方的用戶!
15:49
Or we can monitor doctors prescribing behaviors, passively,
394
934000
3000
我們也可以被動監測醫生開藥的行為，
15:52
and see how the diffusion of innovation with pharmaceuticals
395
937000
3000
以了解對藥品的接受度，
15:55
occurs within [networks of] doctors.
396
940000
2000
是如何在醫生之間擴散的。
15:57
Or again, we can monitor purchasing behavior in people
397
942000
2000
我們也可以監測人們的購買行為，
15:59
and watch how these types of phenomena
398
944000
2000
觀察購買現象是如何
16:01
can diffuse within human populations.
399
946000
3000
在人群中散播的。
16:04
And there are three ways, I think,
400
949000
2000
我認為這些巨量-被動收集所得的資料，
16:06
that these massive-passive data can be used.
401
951000
2000
有三種方式可以利用。
16:08
One is fully passive,
402
953000
2000
一種是完全的被動，
16:10
like I just described --
403
955000
2000
像我剛剛所描述的 －
16:12
as in, for instance, the trucker example,
404
957000
2000
例如卡車司機的例子，
16:14
where we don't actually intervene in the population in any way.
405
959000
2000
我們並不對群體做任何形式的干預。
16:16
One is quasi-active,
406
961000
2000
一種是半主動，
16:18
like the flu example I gave,
407
963000
2000
像是之前流感的例子，
16:20
where we get some people to nominate their friends
408
965000
3000
我們讓某些人舉出他們的朋友，
16:23
and then passively monitor their friends --
409
968000
2000
然後被動的觀察他們的朋友 －
16:25
do they have the flu, or not? -- and then get warning.
410
970000
2000
他們感冒了沒？－ 並據此取得預警。
16:27
Or another example would be,
411
972000
2000
另外一個例子是，
16:29
if you're a phone company, you figure out who's central in the network
412
974000
3000
電信公司可以想辦法找出網路的中心群，
16:32
and you ask those people, "Look, will you just text us your fever every day?
413
977000
3000
問他們，"你能不能每天用簡訊，讓我們知道你發燒了沒？
16:35
Just text us your temperature."
414
980000
2000
只要傳送體溫即可"
16:37
And collect vast amounts of information about people's temperature,
415
982000
3000
然後從中心群體裡，
16:40
but from centrally located individuals.
416
985000
2000
大量收集體溫資料，
16:42
And be able, on a large scale,
417
987000
2000
便能夠用少量的資料輸入，
16:44
to monitor an impending epidemic
418
989000
2000
來進行大規模的監控，
16:46
with very minimal input from people.
419
991000
2000
以預測流感的爆發。
16:48
Or, finally, it can be more fully active --
420
993000
2000
最後是完全主動的方式 －
16:50
as I know subsequent speakers will also talk about today --
421
995000
2000
就我所知下位演講者也會談到－
16:52
where people might globally participate in wikis,
422
997000
2000
現在全世界的人都參與維基百科的編寫、
16:54
or photographing, or monitoring elections,
423
999000
3000
拍攝照片、或是監視選舉，
16:57
and upload information in a way that allows us to pool
424
1002000
2000
人們將資訊上傳，使得我們能夠匯集
16:59
information in order to understand social processes
425
1004000
2000
資訊以了解社會進程，
17:01
and social phenomena.
426
1006000
2000
以及社會現象的產生。
17:03
In fact, the availability of these data, I think,
427
1008000
2000
我認為這些資料的垂手可得，
17:05
heralds a kind of new era
428
1010000
2000
揭示了一個新時代的來臨，
17:07
of what I and others would like to call
429
1012000
2000
我們將之稱作
17:09
"computational social science."
430
1014000
2000
"計算社會科學"。
17:11
It's sort of like when Galileo invented -- or, didn't invent --
431
1016000
3000
有點類似伽利略發明 －或許沒有發明－
17:14
came to use a telescope
432
1019000
2000
望遠鏡的誕生，
17:16
and could see the heavens in a new way,
433
1021000
2000
而可以從全新的角度來觀看天空。
17:18
or Leeuwenhoek became aware of the microscope --
434
1023000
2000
或是雷文霍克發現顯微鏡 －
17:20
or actually invented --
435
1025000
2000
或許是他發明的－
17:22
and could see biology in a new way.
436
1027000
2000
而能夠用新的方式看待生物學。
17:24
But now we have access to these kinds of data
437
1029000
2000
現在我們能夠取得的資料，
17:26
that allow us to understand social processes
438
1031000
2000
能夠讓我們用過去未見的嶄新角度
17:28
and social phenomena
439
1033000
2000
了解社會的進程，
17:30
in an entirely new way that was never before possible.
440
1035000
3000
以及其中發生的現象。
17:33
And with this science, we can
441
1038000
2000
有了這樣的科學，
17:35
understand how exactly
442
1040000
2000
我們就能夠了解
17:37
the whole comes to be greater
443
1042000
2000
群體的綜效，是如何優於
17:39
than the sum of its parts.
444
1044000
2000
單純個體的加總。
17:41
And actually, we can use these insights
445
1046000
2000
我們也能運用這些理解，
17:43
to improve society and improve human well-being.
446
1048000
3000
來增進社會以及人類的福祉。
17:46
Thank you.
447
1051000
2000
謝謝。
ABOUT THE SPEAKER
Nicholas Christakis - Physician, social scientistNicholas Christakis explores how the large-scale, face-to-face social networks in which we are embedded affect our lives, and what we can do to take advantage of this fact.
Why you should listen
People aren't merely social animals in the usual sense, for we don't just live in groups. We live in networks -- and we have done so ever since we emerged from the African savannah. Via intricately branching paths tracing out cascading family connections, friendship ties, and work relationships, we are interconnected to hundreds or even thousands of specific people, most of whom we do not know. We affect them and they affect us.
Nicholas Christakis' work examines the biological, psychological, sociological, and mathematical rules that govern how we form these social networks, and the rules that govern how they shape our lives. His work shows how phenomena as diverse as obesity, smoking, emotions, ideas, germs, and altruism can spread through our social ties, and how genes can partially underlie our creation of social ties to begin with. His work also sheds light on how we might take advantage of an understanding of social networks to make the world a better place.
At Yale, Christakis is a Professor of Social and Natural Science, and he directs a diverse research group in the field of biosocial science, primarily investigating social networks. His popular undergraduate course "Health of the Public" is available as a podcast. His book, Connected, co-authored with James H. Fowler, appeared in 2009, and has been translated into 20 languages. In 2009, he was named by Time magazine to its annual list of the 100 most influential people in the world, and also, in 2009 and 2010, by Foreign Policy magazine to its list of 100 top global thinkers
Nicholas Christakis | Speaker | TED.com