Kalika Bali: The giant leaps in language technology -- and who's left behind | TED

54,150 views

2021-04-26 ใƒป TED


New videos

Kalika Bali: The giant leaps in language technology -- and who's left behind | TED

54,150 views ใƒป 2021-04-26

TED


์•„๋ž˜ ์˜๋ฌธ์ž๋ง‰์„ ๋”๋ธ”ํด๋ฆญํ•˜์‹œ๋ฉด ์˜์ƒ์ด ์žฌ์ƒ๋ฉ๋‹ˆ๋‹ค.

00:00
Transcriber:
0
0
7000
๋ฒˆ์—ญ: Donghyun Oh ๊ฒ€ํ† : DK Kim
00:12
I'm Kalika Bali, I'm a linguist by training
1
12430
2800
์ €๋Š” ์นผ๋ฆฌ์นด ๋ฐœ๋ฆฌ์ž…๋‹ˆ๋‹ค.
์–ธ์–ดํ•™์„ ์ „๊ณตํ–ˆ๊ณ  ์ง์—…์€ ๊ธฐ์ˆ  ์ „๋ฌธ๊ฐ€์ž…๋‹ˆ๋‹ค.
00:15
and a technologist by profession,
2
15263
2334
00:17
I have worked in academia,
3
17630
1667
์ €๋Š” ํ•™๊ณ„, ์Šคํƒ€ํŠธ์—…,
00:19
in startups, in small companies and multinationals for over two decades,
4
19330
5267
์ค‘์†Œ๊ธฐ์—…๊ณผ ๋‹ค๊ตญ์  ๊ธฐ์—…๋“ค์—์„œ ์ด์‹ญ ๋…„ ์ด์ƒ ์ผํ–ˆ์Šต๋‹ˆ๋‹ค.
00:24
doing research in and building language technology systems.
5
24630
3733
์–ธ์–ด ๊ธฐ์ˆ  ์‹œ์Šคํ…œ๊ณผ ๊ด€๋ จํ•˜์—ฌ ์—ฐ๊ตฌ์™€ ์„ค๊ณ„๋ฅผ ๋งก์•˜์Šต๋‹ˆ๋‹ค.
00:28
My dream is to see technology work across the language barrier.
6
28363
4667
์ œ ๊ฟˆ์€ ๊ธฐ์ˆ ์ด ์–ธ์–ด ์žฅ๋ฒฝ์„ ์ดˆ์›”ํ•ด ์ž‘๋™ํ•˜๋Š” ๋ชจ์Šต์„ ๋ณด๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
00:33
As a researcher at Microsoft Research Labs India
7
33030
3100
์ธ๋„ ๋งˆ์ดํฌ๋กœ์†Œํ”„ํŠธ ์—ฐ๊ตฌ์†Œ ์—ฐ๊ตฌ์›์œผ๋กœ
00:36
I work in the field of language technology and speech technology.
8
36163
5534
์–ธ์–ด ๊ธฐ์ˆ  ๋ถ„์•ผ์™€ ์Œ์„ฑ ๊ธฐ์ˆ  ๋ถ„์•ผ์—์„œ ์ผํ•ฉ๋‹ˆ๋‹ค.
00:41
And I worry about how can we make technology accessible
9
41697
3600
์ œ ๊ณ ๋ฏผ์€ ์–ด๋–ป๊ฒŒ ๊ธฐ์ˆ ์„ ๋ชจ๋‘๊ฐ€ ์ ‘๊ทผ ๊ฐ€๋Šฅํ•˜๊ฒŒ ๋งŒ๋“ค๊นŒ์ž…๋‹ˆ๋‹ค.
00:45
to people across the board,
10
45330
2200
00:47
you know, irrespective of the language that they speak.
11
47530
2967
์‚ฌ์šฉํ•˜๋Š” ์–ธ์–ด์™€ ์ƒ๊ด€์—†์ด ๋ง์ด์ฃ .
00:51
So natural language processing,
12
51663
1600
์ž์—ฐ์–ด ์ฒ˜๋ฆฌ, ์ธ๊ณต์ง€๋Šฅ, ์Œ์„ฑ ๊ธฐ์ˆ ,
00:53
artificial intelligence, speech technology,
13
53297
2066
00:55
these are very big words, they are buzzwords right now.
14
55363
2600
์ด ๋‹จ์–ด๋“ค์€ ๋ชจ๋‘ ๋ฒ”์œ„๊ฐ€ ๋„“๊ณ , ํ˜„์žฌ ์œ ํ–‰ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
00:57
Everybody is talking about what exactly is NLP or natural language processing.
15
57997
5533
๋ชจ๋‘๊ฐ€ NLP, ์ฆ‰ ์ž์—ฐ์–ด ์ฒ˜๋ฆฌ๊ฐ€ ์ •ํ™•ํžˆ ๋ฌด์—‡์ธ์ง€์— ๋Œ€ํ•ด ์–˜๊ธฐํ•ฉ๋‹ˆ๋‹ค.
01:03
So in very simple terms,
16
63530
1533
๊ฐ„๋‹จํžˆ ๋งํ•˜์ž๋ฉด,
01:05
this is the part of computer science engineering
17
65063
3367
NLP๋Š” ์ปดํ“จํ„ฐ๊ณผํ•™๊ธฐ์ˆ  ๋ถ„์•ผ ์ค‘
01:08
that makes machines process,
18
68463
2834
๊ธฐ๊ณ„์—๊ฒŒ ์ž์—ฐ์–ด๋ฅผ ์ฒ˜๋ฆฌํ•˜๊ณ ,
01:11
understand and generate natural language,
19
71330
2967
์ดํ•ดํ•˜๊ณ , ์ƒ์„ฑํ•˜๋„๋ก ํ•˜๋Š” ๋ถ„์•ผ์ž…๋‹ˆ๋‹ค.
01:14
which is the language that humans speak.
20
74330
2233
์ž์—ฐ์–ด๋Š” ์ธ๊ฐ„์ด ์‚ฌ์šฉํ•˜๋Š” ์–ธ์–ด์ž…๋‹ˆ๋‹ค.
01:17
When you are interacting with a bot trying to book your train tickets
21
77697
4700
์ž๋™ ์‘๋‹ต ๋ฌธ์ž๋ฅผ ์‚ฌ์šฉํ•ด์„œ
๊ธฐ์ฐจํ‘œ๋‚˜ ๋น„ํ–‰๊ธฐ ํ‘œ๋ฅผ ์˜ˆ๋งคํ•  ๋•Œ๋‚˜,
01:22
or flight tickets,
22
82430
1267
01:23
when you are speaking to a voice-based digital assistant in your phone,
23
83697
4833
ํœด๋Œ€์ „ํ™” ์† ๊ฐ€์ƒ ๋น„์„œ์™€ ๋Œ€ํ™”๋ฅผ ํ•  ๋•Œ
01:28
it's natural language processing
24
88530
1600
์ด๋ฅผ ํ•ด๋‚ด๋Š” ๋ชจ๋“  ๊ธฐ์ˆ ์˜ ๋ฐ”ํƒ•์ด ๋ฐ”๋กœ ์ž์—ฐ์–ด ์ฒ˜๋ฆฌ์ž…๋‹ˆ๋‹ค.
01:30
that underpins the entire technology that makes that work.
25
90163
3534
01:34
But how does this work?
26
94797
1266
๊ทธ๋Ÿฌ๋ฉด NLP๋Š” ์–ด๋–ป๊ฒŒ ์ž‘๋™ํ• ๊นŒ์š”?
01:36
How does NLP work?
27
96063
1567
01:37
In a very, very basic way,
28
97663
3900
์ •๋ง ์•„์ฃผ ๊ฐ„๋‹จํžˆ ๋งํ•˜๋ฉด,
01:41
it's about data.
29
101563
1800
์ž๋ฃŒ๊ฐ€ ์ค‘์š”ํ•ฉ๋‹ˆ๋‹ค.
01:43
So a huge amount of data of how actually humans use language
30
103363
6334
์‚ฌ๋žŒ๋“ค์ด ์‹ค์ œ๋กœ ์–ด๋–ป๊ฒŒ ์–ธ์–ด๋ฅผ ์‚ฌ์šฉํ•˜๋Š”์ง€์— ๋Œ€ํ•œ ๋ฐฉ๋Œ€ํ•œ ์ž๋ฃŒ๊ฐ€
01:49
is then processed by certain algorithms and techniques
31
109697
5166
ํŠน์ •ํ•œ ์•Œ๊ณ ๋ฆฌ์ฆ˜๊ณผ ๊ธฐ์ˆ ๋กœ ์ฒ˜๋ฆฌ๋˜๊ณ 
01:54
that make the machines learn the patterns
32
114863
2767
์ด๋ฅผ ํ†ตํ•ด ๊ธฐ๊ณ„๊ฐ€ ์ธ๊ฐ„์˜ ์ž์—ฐ์–ด ์‚ฌ์šฉ ์–‘์ƒ์„ ์ตํž™๋‹ˆ๋‹ค.
01:57
of natural language of humans, right?
33
117663
3700
02:01
These days, another buzzword that you hear a lot about is deep neural networks.
34
121363
5367
์š”์ฆ˜ ๋งŽ์ด ์–ธ๊ธ‰๋˜๊ณ  ์žˆ๋Š” ๋˜ ๋‹ค๋ฅธ ์œ ํ–‰์–ด๋Š” ์‹ฌ์ธต ์‹ ๊ฒฝ๋ง์ž…๋‹ˆ๋‹ค.
02:06
And these are the advanced techniques
35
126763
2467
์ด๋Ÿฐ ์„ ๋„ ๊ธฐ์ˆ ๋“ค์€
02:09
that underpin a lot of the NLP stuff that happens right now.
36
129263
4134
ํ˜„์žฌ ์ง„ํ–‰ ์ค‘์ธ NLP์˜ ๋งŽ์€ ๋ถ€๋ถ„์„ ๋ฐ‘๋ฐ›์นจํ•ฉ๋‹ˆ๋‹ค.
02:13
And I will not go into the details of how that works,
37
133430
3333
์ด๋–ค ๋ฐฉ์‹์œผ๋กœ ์ž‘๋™ํ•˜๋Š”์ง€ ์ž์„ธํžˆ ๋ง์”€๋“œ๋ฆฌ์ง€๋Š” ์•Š๊ฒ ์ง€๋งŒ,
02:16
but the thing that you really have to understand and keep in mind
38
136797
3400
์ •๋ง ์ดํ•ดํ•˜๊ณ  ๊ธฐ์–ตํ•˜์…”์•ผ ํ•  ๊ฒƒ์€
02:20
is that all of this requires a humungous amount of data,
39
140197
5000
์ด ๋ชจ๋“  ๊ณผ์ •์— ์—„์ฒญ๋‚œ ์ž๋ฃŒ๊ฐ€ ํ•„์š”ํ•˜๋‹ค๋Š” ์ ์ž…๋‹ˆ๋‹ค.
02:25
natural language data.
40
145197
1666
์ž์—ฐ์–ด ์ž๋ฃŒ ๋ง์ž…๋‹ˆ๋‹ค.
02:26
If you want a speech system to converse with you in Gujarati,
41
146863
5267
๊ตฌ์ž๋ผํŠธ์–ด๋กœ ์†Œํ†ตํ•  ์ˆ˜ ์žˆ๋Š” ์–ธ์–ด ์ฒด๊ณ„๋ฅผ ์›ํ•˜์‹ ๋‹ค๋ฉด,
02:32
the first thing you require
42
152163
1367
์ œ์ผ ๋จผ์ € ํ•„์š”ํ•œ ๊ฒƒ์€ ์ˆ˜๋งŽ์€ ๊ตฌ์ž๋ผํŠธ ์‚ฌ๋žŒ๋“ค์ด
02:33
is a lot of data of Gujarati people speaking to each other
43
153530
4533
๊ทธ๋“ค์˜ ์–ธ์–ด๋กœ ์„œ๋กœ ๋Œ€ํ™”ํ•˜๋Š” ์ž๋ฃŒ์ž…๋‹ˆ๋‹ค.
02:38
in their own language.
44
158063
1867
02:41
So 2017, Microsoft came up with a speech recognition system
45
161663
4734
2017๋…„, ๋งˆ์ดํฌ๋กœ์†Œํ”„ํŠธ๋Š” ์Œ์„ฑ ์ธ์‹ ์ฒด๊ณ„๋ฅผ ๋งŒ๋“ค์—ˆ๋Š”๋ฐ
02:46
which was able to transcribe speech into text
46
166430
3600
์Œ์„ฑ์„ ๋ฌธ์ž๋กœ ๋ณ€ํ™˜ํ•˜๋Š” ๋ฐ์— ์‚ฌ๋žŒ๋ณด๋‹ค ๋” ์šฐ์ˆ˜ํ–ˆ์Šต๋‹ˆ๋‹ค.
02:50
better than a human did.
47
170030
1733
02:52
And this system was trained
48
172563
3167
์ด ์ฒด๊ณ„๋Š” 2์–ต ๊ฑด์˜ ๋ณ€ํ™˜ ๋‹จ์–ด๋กœ ํ•™์Šตํ–ˆ์Šต๋‹ˆ๋‹ค.
02:55
on 200 million transcribed words.
49
175763
2667
02:58
In 2018, an English-Chinese machine translation system
50
178463
3967
2018๋…„์—, ์˜์–ด-์ค‘๊ตญ์–ด ๊ธฐ๊ณ„ ๋ฒˆ์—ญ ์ฒด๊ณ„๋Š”
03:02
was able to translate from English to Chinese
51
182463
2800
์˜์–ด์—์„œ ์ค‘๊ตญ์–ด๋กœ ๋ฒˆ์—ญ์„ ์ด์ค‘ ์–ธ์–ด๋ฅผ ์‚ฌ์šฉํ•˜๋Š” ์‚ฌ๋žŒ๋งŒํผ ํ•  ์ˆ˜ ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.
03:05
as well as any human bilingual could.
52
185297
2766
03:08
And this was trained on 18 million bilingual sentence pairs.
53
188063
4934
์ด ์ฒด๊ณ„๋Š” ์ด์ค‘ ์–ธ์–ด ๋ฌธ์žฅ ์ฒœํŒ”๋ฐฑ๋งŒ ์Œ์„ ํ•™์Šตํ–ˆ์Šต๋‹ˆ๋‹ค.
03:14
This is a very, very exciting time in natural language processing
54
194330
4433
์ง€๊ธˆ์€ ์ž์—ฐ์–ด ์ฒ˜๋ฆฌ๋‚˜ ๊ด€๋ จ ๊ธฐ์ˆ ์—์„œ
03:18
and in technology as such.
55
198797
1466
์•„์ฃผ ๋งค์šฐ ํฅ๋ฏธ์ง„์ง„ํ•œ ์‹œ๊ฐ„์ž…๋‹ˆ๋‹ค.
03:20
You know, we are seeing science fiction, which we had read about and watched,
56
200297
4466
์šฐ๋ฆฌ๊ฐ€ ์ฝ๊ณ  ๋ด ์™”๋˜ ๊ณต์ƒ ๊ณผํ•™์ด
03:24
kind of come true in front of our own eyes.
57
204797
2833
ํ˜„์‹ค์ด ๋˜๋Š” ๊ฒƒ์„ ๋ˆˆ์œผ๋กœ ๋ณด๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
03:27
We are making giant leaps in technical advancement.
58
207663
4467
์šฐ๋ฆฌ๋Š” ๊ธฐ์ˆ ์  ๋ฐœ์ „์—์„œ ๊ฑฐ๋Œ€ํ•œ ๋„์•ฝ์„ ํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
03:32
But these giant leaps are limited to very few languages.
59
212163
6234
๊ทธ๋Ÿฌ๋‚˜ ์ด ๊ฑฐ๋Œ€ํ•œ ๋„์•ฝ์€ ์•„์ฃผ ์†Œ์ˆ˜์˜ ์–ธ์–ด๋“ค์— ์ œํ•œ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค.
03:38
So Monojit Choudhury,
60
218430
1300
์ œ ์ข‹์€ ์นœ๊ตฌ์ด์ž ๋™๋ฃŒ์ธ ๋ชจ๋…ธ์ง€ํŠธ ์ดˆ๋”๋ฆฌ๋Š”
03:39
who's like a very good friend of mine
61
219763
1867
03:41
and a colleague,
62
221663
2134
03:43
he has studied this in some detail
63
223830
1733
์ด ์ƒํ™ฉ์„ ์ž์„ธํžˆ ์—ฐ๊ตฌํ•˜์˜€๊ณ ,
03:45
and he has looked at resource distribution across languages in the world.
64
225563
4067
์„ธ๊ณ„ ์–ธ์–ด๋“ค์— ๋Œ์•„๊ฐ€๋Š” ์ž์› ๋ถ„๋ฐฐ๋ฅผ ์‚ดํŽด๋ณด์•˜์Šต๋‹ˆ๋‹ค.
03:49
And he says that these follow what is called a power-law distribution,
65
229663
4100
๊ทธ๋Š” ๋ถ„๋ฐฐ๊ฐ€ ๋ฉฑํ•จ์ˆ˜ ๋ถ„ํฌ๋ผ ๋ถ€๋ฅด๋Š” ๊ฒƒ์„ ๋”ฐ๋ฅธ๋‹ค๊ณ  ๋งํ•ฉ๋‹ˆ๋‹ค.
03:53
which essentially means that there are four languages,
66
233797
2766
๊ธฐ๋ณธ์ ์œผ๋กœ ๋„ค ๊ฐ€์ง€ ์–ธ์–ด,
03:56
Arabic, Chinese, English and Spanish,
67
236563
2700
์•„๋ž์–ด, ์ค‘๊ตญ์–ด, ์˜์–ด์™€ ์ŠคํŽ˜์ธ์–ด๊ฐ€
03:59
which have the maximum amount of resources available.
68
239297
3766
์‚ฌ์šฉํ•  ์ˆ˜ ์žˆ๋Š” ์ž์›์˜ ๊ฑฐ์˜ ๋ชจ๋‘๋ฅผ ์ฐจ์ง€ํ•ฉ๋‹ˆ๋‹ค.
04:03
There are another handful of languages which can also benefit from, you know,
69
243063
5167
ํ˜„์žฌ ์‚ฌ์šฉ ๊ฐ€๋Šฅํ•œ ์ž์›๊ณผ ๊ธฐ์ˆ ์„ ๋ˆ„๋ฆฌ๋Š”
๋‹ค๋ฅธ ์–ธ์–ด๋“ค์ด ์กฐ๊ธˆ ๋” ์žˆ์Šต๋‹ˆ๋‹ค.
04:08
the resources and the technology that's available right now.
70
248263
3834
04:12
But there are 90 percent of the world's languages
71
252130
4833
๊ทธ๋Ÿฌ๋‚˜ ํ˜„์žฌ ์„ธ๊ณ„ ์–ธ์–ด์˜ 90%๋Š”
04:16
which have no resources
72
256997
1800
04:18
or very little resources available.
73
258830
2067
์•„์ฃผ ์ ์€ ์ž์›๋งŒ์„ ์‚ฌ์šฉํ•  ์ˆ˜ ์žˆ๊ฑฐ๋‚˜ ์‚ฌ์šฉํ•  ์ž์›์ด ์—†์Šต๋‹ˆ๋‹ค.
04:20
This revolution that we are talking about
74
260930
2667
์šฐ๋ฆฌ๊ฐ€ ๋งํ•˜๊ณ  ์žˆ๋Š” ์ด ํ˜๋ช…์€
04:23
has essentially bypassed 5,000 languages of the world.
75
263630
4100
๊ทผ๋ณธ์ ์œผ๋กœ ์„ธ๊ณ„์˜ ์–ธ์–ด 5์ฒœ ๊ฐœ๋ฅผ ๋ฌด์‹œํ•˜๋Š” ์…ˆ์ž…๋‹ˆ๋‹ค.
04:27
Now, what this means is that resource-rich languages
76
267763
2534
์ด ํ˜„์ƒ์ด ์˜๋ฏธํ•˜๋Š” ๋ฐ”๋Š”, ์ž์›์ด ํ’๋ถ€ํ•œ ์–ธ์–ด๋“ค์€
04:30
have technologies built for them,
77
270330
1800
ํ•ด๋‹น ์–ธ์–ด๋“ค์„ ์œ„ํ•œ ๊ธฐ์ˆ ์ด ๊ตฌ์ถ•๋˜์–ด ์žˆ์œผ๋ฏ€๋กœ
04:32
so researchers and technologists get attracted towards them.
78
272163
3267
์—ฐ๊ตฌ์ž๋“ค๊ณผ ๊ธฐ์ˆ  ์ „๋ฌธ๊ฐ€๋“ค์ด ๋ชจ์ž…๋‹ˆ๋‹ค.
04:35
They build more technologies for them. They create more resources.
79
275463
3500
๊ทธ๋“ค์€ ํ•ด๋‹น ์–ธ์–ด๋“ค์— ๋” ๋งŽ์€ ๊ธฐ์ˆ ์„ ์Œ“๊ณ  ๋” ๋งŽ์€ ์ž์›์„ ๋งŒ๋“ญ๋‹ˆ๋‹ค.
04:38
So it's like a rich getting richer kind of a cycle.
80
278997
2800
๋ถ€์ž๊ฐ€ ๋” ๋ถ€์ž๊ฐ€ ๋˜๋Š” ์ˆœํ™˜์ž…๋‹ˆ๋‹ค.
04:41
And the resource-poor languages stay poor,
81
281830
2400
์ž์›์ด ์—†๋Š” ์–ธ์–ด๋“ค์€ ๊ณ„์† ๋นˆ๊ณคํ•˜๊ณ ,
04:44
there's no technology for them, nobody works for them.
82
284263
2600
๊ธฐ์ˆ ๋„ ์—†์œผ๋ฉฐ ์•„๋ฌด๋„ ๊ทธ ์–ธ์–ด๋ฅผ ์—ฐ๊ตฌํ•˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
04:46
And this divide, digital divide between languages
83
286863
3400
์–ธ์–ด ๊ฐ„์˜ ๋””์ง€ํ„ธ ๊ฒฉ์ฐจ๋Š” ์ ์  ํ™•๋Œ€๋˜๊ณ 
04:50
is ever-expanding
84
290297
1500
04:51
and by implication also the divide between the communities
85
291830
4633
๊ฒฐ๊ณผ์ ์œผ๋กœ ์–ธ์–ด๋ฅผ ์‚ฌ์šฉํ•˜๋Š”
04:56
that speak these languages is expanding.
86
296497
2500
๊ฐ ์‚ฌํšŒ ๊ฐ„์˜ ๊ฒฉ์ฐจ ๋˜ํ•œ ํ™•๋Œ€๋˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
05:00
So in Microsoft, in Project Ellora, we aim to bridge this gap.
87
300763
4767
๋งˆ์ดํฌ๋กœ์†Œํ”„ํŠธ์˜ ์—˜๋กœ๋ผ ์‚ฌ์—…์€ ์ด ๊ฐ„๊ทน์„ ๋ฉ”์šฐ๊ณ ์ž ํ•ฉ๋‹ˆ๋‹ค.
05:06
We are trying to see how can we create more data by innovative methods,
88
306663
5567
ํ˜์‹ ์ ์ธ ๋ฐฉ๋ฒ•๋ก ์„ ๋„์ž…ํ•ด ์–ด๋–ป๊ฒŒ ๋” ๋งŽ์€ ์ž๋ฃŒ๋ฅผ ๋งŒ๋“ค ์ˆ˜ ์žˆ์„์ง€,
05:12
have more techniques to build technology without having a lot of resources,
89
312263
5800
๋งŽ์€ ์ž์›์ด ์—†์ด๋„ ๊ธฐ์ˆ ์„ ๊ตฌ์ถ•ํ•  ๋” ๋งŽ์€ ๋ฐฉ๋ฒ•์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ๋Š”์ง€,
05:18
and what are the applications that can truly benefit these communities.
90
318063
4200
์ง„์ • ์ด ์‚ฌํšŒ์— ๋„์›€์ด ๋  ์ ์šฉ ๋ฐฉ์‹์ด ๋ฌด์—‡์ธ์ง€ ์—ฐ๊ตฌํ•ฉ๋‹ˆ๋‹ค.
05:23
So at the moment, this might seem very theoretical,
91
323463
3334
์ง€๊ธˆ์€ ๋ชจ๋“  ๊ฒƒ์ด ๋ฌด์ฒ™ ์ด๋ก ์ ์œผ๋กœ ๋ณด์ž…๋‹ˆ๋‹ค.
05:26
like what is he talking about, data and techniques and technology.
92
326830
3133
์ž๋ฃŒ์™€ ๋ฐฉ๋ฒ•๋ก ๊ณผ ๊ธฐ์ˆ ์„ ์–ด๋–ป๊ฒŒ ํ•œ๋‹ค๋Š” ๊ฒƒ์ผ๊นŒ์š”.
05:29
So let me give you a very concrete example here.
93
329997
3066
์—ฌ๊ธฐ ๊ตฌ์ฒด์ ์ธ ์˜ˆ์‹œ๋ฅผ ๋“œ๋ฆฌ๊ฒ ์Šต๋‹ˆ๋‹ค.
05:33
I'm a linguist at heart, I love languages, and that's what I love talking about.
94
333063
5300
์ €๋Š” ๋งˆ์Œ์†์œผ๋กœ๋Š” ์–ธ์–ดํ•™์ž์ž…๋‹ˆ๋‹ค.
์–ธ์–ด๋ฅผ ์‚ฌ๋ž‘ํ•˜๊ณ  ์–ธ์–ด์— ๋Œ€ํ•ด ๋งํ•˜๋Š” ๊ฒƒ์„ ์ข‹์•„ํ•ฉ๋‹ˆ๋‹ค.
05:38
So let me tell you about a language that many of you might not know about.
95
338363
4367
์—ฌ๋Ÿฌ๋ถ„ ์ค‘ ๋Œ€๋ถ€๋ถ„์ด ์ž˜ ๋ชจ๋ฅด์‹ค ์–ธ์–ด์— ๋Œ€ํ•ด ๋ง์”€๋“œ๋ฆฌ๊ณ ์ž ํ•ฉ๋‹ˆ๋‹ค.
05:42
Gondi.
96
342763
1267
๋ฐ”๋กœ ๊ณค๋“œ์–ด์ž…๋‹ˆ๋‹ค.
05:44
Gondi is a South-Central Dravidian language.
97
344030
2700
๊ณค๋“œ์–ด๋Š” ์ค‘๋‚จ๋ถ€ ๋“œ๋ผ๋น„๋‹ค์–ด์— ์†ํ•˜๋Š” ์–ธ์–ด์ž…๋‹ˆ๋‹ค.
05:46
It is spoken by three million people in five states of India.
98
346763
4434
์ธ๋„ 5๊ฐœ์ฃผ์—์„œ 3๋ฐฑ๋งŒ ๋ช…์ด ์‚ฌ์šฉํ•˜๋Š” ์–ธ์–ด์ž…๋‹ˆ๋‹ค.
05:51
And to put this in some kind of perspective,
99
351197
3000
์ด๋ฅผ ๊ฐ๊ด€์ ์ธ ์‹œ๊ฐ์œผ๋กœ ๋ฐ”๋ผ๋ณธ๋‹ค๋ฉด,
05:54
Norwegian is spoken by five million people
100
354197
2833
๋…ธ๋ฅด์›จ์ด์–ด๋Š” 5๋ฐฑ๋งŒ ๋ช…์ด ์‚ฌ์šฉํ•˜๋ฉฐ
05:57
and Welsh by a little under a million.
101
357030
2933
์›จ์ผ์Šค์–ด๋Š” ๋ฐฑ๋งŒ ๋ช…์ด ์กฐ๊ธˆ ์•ˆ ๋ฉ๋‹ˆ๋‹ค.
05:59
So Gondi is actually a pretty robust and pretty large community
102
359997
6200
๊ทธ๋Ÿฌ๋‹ˆ ๊ณค๋“œ์–ด๋Š” ์‚ฌ์‹ค
์ธ๋„ ๊ณค๋“œ์กฑ๋“ค์˜ ๊ฝค๋‚˜ ํŠผํŠผํ•˜๊ณ 
06:06
of the Gond tribals in India.
103
366197
2900
์•„์ฃผ ๊ฑฐ๋Œ€ํ•œ ์‚ฌํšŒ์ธ ์…ˆ์ž…๋‹ˆ๋‹ค.
06:09
But by UNESCO's Atlas of Languages in Danger,
104
369130
5067
๊ทธ๋Ÿฌ๋‚˜ ์œ ๋„ค์Šค์ฝ”์˜ ์†Œ๋ฉธ ์œ„๊ธฐ ์–ธ์–ด ์ง€๋„์—์„œ
06:14
Gondi is designated vulnerable status.
105
374197
4366
๊ณค๋“œ์–ด๋Š” ์œ„ํ—˜ ๋“ฑ๊ธ‰์„ ๋ฐ›์•˜์Šต๋‹ˆ๋‹ค.
06:19
CGNet Swara is an NGO that provides a citizen journalism portal
106
379030
4533
CG๋„ท ์Šค์™€๋ผ๋Š” ๊ณค๋“œ์–ด ์‚ฌํšŒ๋ฅผ ์œ„ํ•œ
06:23
for the Gond community
107
383563
1867
์‹œ๋ฏผ ์–ธ๋ก  ๊ด€๋ฌธ์„ ์ œ๊ณตํ•˜๋Š” NGO๋กœ์„œ
06:25
by making local stories accessible through mobile phones.
108
385463
4234
์ง€์—ญ ๊ธฐ์‚ฌ๊ฑฐ๋ฆฌ๋“ค์„ ํœด๋Œ€์ „ํ™”๋กœ ์ ‘ํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•ด์ค๋‹ˆ๋‹ค.
06:29
There's absolutely no tech support for Gondi.
109
389697
2933
๊ณค๋“œ์–ด์—๋Š” ๊ธฐ์ˆ ์  ์ง€์›์ด ์ „ํ˜€ ์—†์Šต๋‹ˆ๋‹ค.
06:32
There is no data available for Gondi, no resources available for Gondi.
110
392663
4800
๊ณค๋“œ์–ด ์ž๋ฃŒ๋„ ์—†๊ณ  ์‚ฌ์šฉ ๊ฐ€๋Šฅํ•œ ์ž์›๋„ ์—†์Šต๋‹ˆ๋‹ค.
06:37
So all content that is created, moderated and edited is done manually.
111
397497
5066
๊ทธ๋ž˜์„œ ๋ชจ๋“  ๊ธฐ์‚ฌ์˜ ์ƒ์„ฑ, ์กฐ์ ˆ, ํŽธ์ง‘์„ ์‚ฌ๋žŒ์˜ ์†์œผ๋กœ ํ•ฉ๋‹ˆ๋‹ค.
06:42
Now, under Project Ellora,
112
402563
2067
ํ˜„์žฌ, ์—˜๋กœ๋ผ ์‚ฌ์—…์„ ํ†ตํ•ด
06:44
what we did was that we brought together all the stakeholders,
113
404663
2967
์ดํ•ด ๊ด€๊ณ„์ž๋“ค์„ ๋ชจ๋‘ ๋ชจ์•˜์Šต๋‹ˆ๋‹ค.
06:47
an NGOs like CGNet Swara,
114
407663
1800
CG๋„ท ์Šค์™€๋ผ ๊ฐ™์€ NGO๋“ค,
06:49
and academic institutions, like IIIT Naya Raipur,
115
409497
3366
IIIT ๋‚˜์•ผ ๋ผ์ดํ‘ธ๋ฅด ๊ฐ™์€ ํ•™์ˆ  ๊ธฐ๊ด€๋“ค๋„ ๋ชจ์•˜์œผ๋ฉฐ,
06:52
a not-for-profit children's book publisher,
116
412863
2200
ํ”„๋ผํƒ ์„œ์  ๊ฐ™์€ ๋น„์˜๋ฆฌ ์•„๋™ ์„œ์  ์ถœํŒ์‚ฌ๋„ ๋ชจ์œผ๊ณ ,
06:55
like Pratham Books,
117
415063
1300
06:56
and most importantly, the speakers of the community.
118
416363
2434
๊ฐ€์žฅ ์ค‘์š”ํ•œ ์–ธ์–ด ์‚ฌ์šฉ์ž๋„ ๋ชจ์•˜์Šต๋‹ˆ๋‹ค.
06:58
The Gond tribals themselves participated in this activity
119
418830
4933
๊ณค๋“œ์–ด ๋ถ€์กฑ๋“ค์ด ์ด ํ™œ๋™์— ์ง์ ‘ ์ฐธ์—ฌํ–ˆ์œผ๋ฉฐ
07:03
and for the first time edited and translated childrenโ€™s books in Gondi.
120
423797
5633
๊ณค๋“œ์–ด๋กœ ๋œ ์•„๋™์šฉ ์„œ์ ๋“ค์„ ์ตœ์ดˆ๋กœ ํŽธ์ง‘ํ•˜๊ณ  ๋ฒˆ์—ญํ–ˆ์Šต๋‹ˆ๋‹ค.
07:09
We were able to put out 200 books for the very first time in Gondi,
121
429463
5334
๊ณค๋“œ์–ด๋กœ ๋œ ์ฑ… 200๊ถŒ์„ ์ตœ์ดˆ๋กœ ๋ฐœํ–‰ํ•ด
07:14
so that the children had access to stories and books in their own language.
122
434830
4700
์•„๋™๋“ค์ด ๊ทธ๋“ค์˜ ์–ธ์–ด๋กœ ๋œ ์ฑ…๊ณผ ์ด์•ผ๊ธฐ๋ฅผ ๋ณผ ์ˆ˜ ์žˆ๋„๋ก ํ–ˆ์Šต๋‹ˆ๋‹ค.
07:19
Another extension of this was Adivasi Radio,
123
439530
2267
์ด ํ™œ๋™์˜ ๋˜ ๋‹ค๋ฅธ ๊ฐˆ๋ž˜๋Š” ์•„๋””๋ฐ”์‹œ ๋ผ๋””์˜ค์˜€๋Š”๋ฐ,
07:21
which was like an app that we built and developed in Microsoft Research,
124
441830
4033
๋งˆ์ดํฌ๋กœ์†Œํ”„ํŠธ ์—ฐ๊ตฌ์†Œ๊ฐ€ ์„ค๊ณ„ํ•˜๊ณ  ๊ฐœ๋ฐœํ•ด์„œ
07:25
and then put out there, along with our stakeholders,
125
445863
4567
์ดํ•ด๊ด€๊ณ„์ž๋“ค๊ณผ ํ•จ๊ป˜ ๋งŒ๋“  ์•ฑ์ž…๋‹ˆ๋‹ค.
07:30
which takes a Hindi text-to-speech system
126
450463
3067
์ด ์•ฑ์€ ํžŒ๋””์–ด ๋ฌธ์ž-์Œ์„ฑ ๋ณ€ํ™˜ ์ฒด๊ณ„๋ฅผ ์ด์šฉํ•ด
07:33
and allows it to read out news and articles provided by CGNet Swara
127
453530
6400
CG๋„ท ์Šค์™€๋ผ๊ฐ€ ์ œ๊ณตํ•˜๋Š”
๋‰ด์Šค์™€ ๊ธฐ์‚ฌ๋“ค์„ ๊ณค๋“œ์–ด๋กœ ์ฝ์–ด์ค๋‹ˆ๋‹ค.
07:39
in Gondi language.
128
459963
2434
07:42
Users can now use this app to read,
129
462430
2767
์‚ฌ์šฉ์ž๋“ค์€ ์ด ์•ฑ์„ ํ™œ์šฉํ•ด
07:45
watch news and access any information
130
465197
3433
์ž์‹ ์ด ์‚ฌ์šฉํ•˜๋Š” ์–ธ์–ด์˜ ๋ฌธ์ž์™€ ์Œ์„ฑ์œผ๋กœ
07:48
through text and voice in their own language.
131
468663
3900
๋‰ด์Šค๋ฅผ ์ฝ๊ฑฐ๋‚˜ ์‹œ์ฒญํ•˜๊ณ  ์–ด๋–ค ์ •๋ณด๋“  ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
07:52
A very interesting thing is that this app is now being used to translate --
132
472563
3634
์ •๋ง ํฅ๋ฏธ๋กœ์šด ๊ฒƒ์€ ์ด ์•ฑ์ด ์ด์ œ ๋ฒˆ์—ญ์— ์‚ฌ์šฉ๋˜๊ณ  ์žˆ๋‹ค๋Š” ์ ์ž…๋‹ˆ๋‹ค.
07:56
by the community to translate text from Hindi to Gondi.
133
476197
4833
์‚ฌ์šฉ์ž๋“ค์ด ํžŒ๋‘์–ด๋ฅผ ๊ณค๋“œ์–ด๋กœ ๋ฒˆ์—ญํ•˜๋Š” ๋ฐ ์‚ฌ์šฉํ•˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
08:01
Now, what that will result in is a lot of parallel data,
134
481030
3333
์ด๋Š” ์ž๋ฃŒ ์Œ์„ ๋งŽ์ด ๋งŒ๋“ค ๊ฒƒ์ธ๋ฐ,
์ด๊ฒƒ์„ ๋ณ‘๋ ฌ ์ž๋ฃŒ๋ผ๊ณ  ๋ถ€๋ฆ…๋‹ˆ๋‹ค.
08:04
that we call parallel data,
135
484363
1534
08:05
that will allow us to build machine translation systems for Gondi,
136
485930
3500
์ด ์ง€๋ฃŒ๋“ค์€ ๊ณค๋“œ์–ด ๊ธฐ๊ณ„ ๋ฒˆ์—ญ ์ฒด๊ณ„ ๊ตฌ์ถ• ๊ฐ€๋Šฅ์„ฑ์„ ์—ด์–ด์ฃผ๋ฉฐ,
08:09
which will truly open up a window for the Gond community to the world.
137
489463
5900
์ด ์ฒด๊ณ„๋Š” ๊ณค๋“œ์–ด ์‚ฌ์šฉ์ž์™€ ์„ธ๊ณ„๋ฅผ ์—ฐ๊ฒฐํ•˜๋Š” ๊ธฐํšŒ๋ฅผ ์ œ๊ณตํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค.
08:15
And what is even more important is now we know how to do this.
138
495363
3567
๋”์šฑ ๋” ์ค‘์š”ํ•œ ๊ฒƒ์€, ์ด์ œ ๋ฐฉ๋ฒ•๋ก ์„ ์–ป์—ˆ๋‹ค๋Š” ์ ์ž…๋‹ˆ๋‹ค.
08:18
We have the entire pipeline and we can replicate this for any language
139
498963
4600
์ด์ œ ์™„์ „ํ•œ ๋ฐฉ๋ฒ•๋ก ์„ ๊ตฌ์ถ•ํ–ˆ์œผ๋ฉฐ, ์–ด๋–ค ์–ธ์–ด์—๋„ ์ ์šฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
08:23
and any language community
140
503563
1900
๊ณค๋“œ์–ด์™€ ๋น„์Šทํ•œ ์ƒํ™ฉ์— ์žˆ๋Š” ์–ด๋Š ์–ธ์–ด์—๋“  ์ ์šฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
08:25
which is in a similar situation as the Gond tribals.
141
505497
3033
08:29
Also education -- yes, you know, information access -- yes,
142
509630
4833
๋˜ํ•œ ๊ต์œก, ์ •๋ณด ํš๋“์—๋„ ํšจ๊ณผ์ ์ด์ง€๋งŒ
08:34
but what about earning a living?
143
514497
2700
์ƒ๊ณ„ ์œ ์ง€ ๋ฌธ์ œ๋Š” ์–ด๋–จ๊นŒ์š”?
08:37
Right? What about -- how can we make these people earn a living
144
517830
4200
ํ˜„์žฌ ๋ชจ๋‘๊ฐ€ ๋‹น์—ฐํ•˜๋‹ค๊ณ  ์—ฌ๊ธฐ๋Š” ๋””์ง€ํ„ธ ๋„๊ตฌ๋“ค์„ ํ†ตํ•ด
08:42
through the digital tools that all of us just take for granted these days?
145
522030
3867
์ด๋“ค์ด ์ƒ๊ณ„๋ฅผ ์œ ์ง€ํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•˜๋Š” ๋ฐฉ์•ˆ์—” ๋ฌด์—‡์ด ์žˆ์„๊นŒ์š”?
08:45
Vivek Seshadri, who's another researcher at MSR,
146
525930
2533
MSR์—์„œ ์—ฐ๊ตฌํ•˜๋Š” ๋น„๋ฒก ์„ธ์ƒค๋“œ๋ฆฌ์™€
08:48
and his collaborator, Manu Chopra,
147
528497
2100
๋™๋ฃŒ์ธ ๋งˆ๋ˆ„ ์ดˆํ”„๋ผ๋Š”
08:50
they've designed a platform called Karya
148
530630
2500
์นด๋ฆฌ์•ผ๋ผ๋Š” ํ”Œ๋žซํผ์„ ์„ค๊ณ„ํ–ˆ๋Š”๋ฐ
08:53
for providing digital microtasks to the underserved communities.
149
533163
4600
์ด๋Š” ์ง€์›๋ฐ›์ง€ ๋ชปํ—€๋˜ ๊ณต๋™์ฒด์— ๋””์ง€ํ„ธ ์ผ๊ฑฐ๋ฆฌ๋ฅผ ์ œ๊ณตํ•ฉ๋‹ˆ๋‹ค.
08:57
His aim was basically to find a way to provide a means of dignified labor
150
537797
5433
๋ชฉํ‘œ๋Š” ๊ธฐ๋ณธ์ ์œผ๋กœ ์ง€์—ญ ๊ฑฐ์ฃผ๋ฏผ๊ณผ ๋„์‹œ ์† ๋นˆ๊ณค ์ธ๊ตฌ์—๊ฒŒ
09:03
to the populations, the rural populations
151
543263
2034
๊ดœ์ฐฎ์€ ์ˆ˜์ค€์˜ ๋…ธ๋™์„ ์ œ๊ณตํ•  ๋ฐฉ๋ฒ•์„ ์ฐพ๋Š” ๊ฒƒ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
09:05
and the urban poor populations of this country.
152
545330
2300
09:08
They don't have access to all the knowledge
153
548530
3000
๊ทธ๋“ค์€ ์šฐ๋ฆฌ๊ฐ€ ์•„๋ฌด ์ƒ๊ฐ๋„ ์•Š๊ณ  ์“ฐ๋Š”
09:11
to use the digital platforms
154
551530
2567
๋””์ง€ํ„ธ ํ”Œ๋žซํผ์„ ์‚ฌ์šฉํ•˜๋Š” ๋ฐ ํ•„์š”ํ•œ ๋ชจ๋“  ์ง€์‹์—
09:14
that all of us use every day without even thinking, right?
155
554130
4200
์ ‘๊ทผ์กฐ์ฐจ ํ•  ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค.
09:18
But ...
156
558930
1233
๊ทธ๋Ÿฌ๋‚˜โ€ฆ
09:20
Here is a large
157
560863
2434
์—ฌ๊ธฐ ์ผํ•˜๊ณ  ์‹ถ์–ดํ•˜๋Š”
09:23
literate population that wants to work, right,
158
563330
4133
๊ธ€์„ ์ฝ๊ณ  ์“ธ ์ˆ˜ ์žˆ๋Š” ์ˆ˜๋งŽ์€ ์‚ฌ๋žŒ๋“ค์ด ์žˆ๋Š”๋ฐ
์–ด๋–ป๊ฒŒ ์ด๋“ค์—๊ฒŒ ์ผ์„ ์ œ๊ณตํ•  ์ˆ˜ ์žˆ์„๊นŒ์š”?
09:27
and how can we make this possible for them?
159
567497
2766
09:30
So Karya is one such way
160
570297
3500
์นด๋ฆฌ์•ผ๋Š” ์ด๋“ค์ด
๋””์ง€ํ„ธ ์„ธ๊ณ„์— ๋‹ฟ์„ ์ˆ˜ ์žˆ๋Š” ๋ฐฉ๋ฒ•์ด๊ณ 
09:33
through which this population can get on to the digital world
161
573830
4033
09:37
and, you know,
162
577863
1434
๊ทธ๋ฆฌ๊ณ , ๊ทธ๋ฅผ ํ†ตํ•ด์„œ
09:39
through that find work and do tasks that can then earn them money.
163
579330
4433
์ง์—…๊ณผ ์ƒ๊ณ„ ์œ ์ง€ ํ™œ๋™์„ ์ฐพ์„ ์ˆ˜ ์žˆ๋Š” ๋ฐฉ๋ฒ•์ž…๋‹ˆ๋‹ค.
09:43
So we saw this and we thought, oh, this is wonderful.
164
583797
2500
์ €ํฌ๋Š” ์ด๋ฅผ ๋ณด๊ณ  ์ƒ๊ฐํ–ˆ์Šต๋‹ˆ๋‹ค.
์ •๋ง ๋Œ€๋‹จํ•˜๋„ค, ์ด ๋ฐฉ๋ฒ•์„ ์ž๋ฃŒ ์ˆ˜์ง‘์—๋„ ์‚ฌ์šฉํ•  ์ˆ˜ ์žˆ๊ฒ ๋Š”๋ฐ.
09:46
We could probably use this for data collection as well.
165
586330
2600
09:48
So we went to Amale,
166
588963
2000
์ €ํฌ๋Š” ์•„๋ง๋ฆฌ๋กœ ํ–ฅํ—€์Šต๋‹ˆ๋‹ค.
09:50
which is a small village of 200 people
167
590997
3266
๋งˆํ•˜๋ผ์ŠˆํŠธ๋ผ์ฃผ ์™€๋‹ค ์ง€์—ญ์— ์žˆ๋Š” ์ธ๊ตฌ 200๋ช…์˜ ์ž‘์€ ๋งˆ์„์ž…๋‹ˆ๋‹ค.
09:54
in the Wada district of Maharashtra
168
594297
1933
09:56
and decided to use Karya to collect Marathi data.
169
596263
2667
๋งˆ๋ผํ‹ฐ์–ด ์ž๋ฃŒ ์ˆ˜์ง‘์— ์นด๋ฆฌ์•ผ๋ฅผ ์“ฐ๊ธฐ๋กœ ํ–ˆ์Šต๋‹ˆ๋‹ค.
09:58
Now, I know what you are thinking --
170
598963
1900
์—ฌ๋Ÿฌ๋ถ„์˜ ์ƒ๊ฐ์„ ์••๋‹ˆ๋‹ค.
10:00
I'm sure a lot of Marathi speakers also in the audience --
171
600930
2767
์ฒญ์ค‘ ์ค‘์—์„œ๋„ ๋งˆ๋ผํ‹ฐ์–ด๋ฅผ ์“ฐ๋Š” ๋ถ„์ด ๋งŽ์„ ํ…๋ฐ
10:03
that Marathi is not a low-resource language.
172
603697
2400
๋งˆ๋ผํ‹ฐ์–ด๋Š” ์ž์›์ด ๋‚ฎ์€ ์–ธ์–ด๊ฐ€ ์•„๋‹ˆ๋ผ๊ณ  ์ƒ๊ฐํ•˜์‹œ๊ฒ ์ฃ .
10:06
Marathi is definitely a mainstream language of the country.
173
606130
3600
๋งˆ๋ผํ‹ฐ์–ด๋Š” ํ™•์‹คํžˆ ์ธ๋„์˜ ์ฃผ์š” ์–ธ์–ด์ž…๋‹ˆ๋‹ค.
10:09
But as far as language technology is concerned,
174
609763
2500
๊ทธ๋Ÿฌ๋‚˜ ์–ธ์–ด ๊ธฐ์ˆ ๋ฉด์—์„œ ๋ณด๋ฉด ๋งˆ๋ผํ‹ฐ์–ด๋Š” ์ž์›์ด ์ ์€ ์–ธ์–ด์ž…๋‹ˆ๋‹ค.
10:12
Marathi is a low-resource language.
175
612297
2333
10:14
So we went to this village
176
614663
1600
์ €ํฌ๋Š” ์ด ๋งˆ์„๋กœ ๊ฐ€์„œ ์•„์ฃผ ์„ฑ๊ณต์ ์œผ๋กœ ์ž๋ฃŒ๋ฅผ ๋ชจ์•˜์Šต๋‹ˆ๋‹ค.
10:16
and we had a very successful data-collection trip.
177
616297
3800
10:20
And, you know, this village is very remote.
178
620130
3567
๊ทธ๋ฆฌ๊ณ  ์ด ๋งˆ์„์€ ์ •๋ง ์™ธ์ง„ ๊ณณ์— ์žˆ์Šต๋‹ˆ๋‹ค.
10:23
They have no TV, they have no electricity,
179
623697
2700
TV๋„ ์—†๊ณ , ์ „๊ธฐ๋„ ์—†์œผ๋ฉฐ, ํœด๋Œ€์ „ํ™” ์‹ ํ˜ธ๋„ ์žกํžˆ์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
10:26
they have no mobile signal.
180
626430
3833
์–ธ๋•์„ ์˜ฌ๋ผ์„œ ์†์„ ๋ป—์–ด์„œ ์ „ํ™”๊ธฐ๋ฅผ ์ด๋ฆฌ์ €๋ฆฌ ํœ˜์ €์–ด์•ผ
10:30
You have to climb a hill and wave your phone around
181
630297
2633
10:32
if you want to, you know, use your mobile to call anyone.
182
632963
3200
ํœด๋Œ€์ „ํ™”๋ฅผ ์‚ฌ์šฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
10:37
So they gave us all this data.
183
637197
1766
๊ฑฐ๊ธฐ์—์„œ ๋ชจ๋“  ์ž๋ฃŒ๋ฅผ ์–ป์—ˆ๋Š”๋ฐ ๊ทธ๋ณด๋‹ค๋„ ๊ฐ’์ง„ ์‚ถ์˜ ๊ตํ›ˆ์„ ์ฃผ์…จ์Šต๋‹ˆ๋‹ค.
10:38
But more than that, they gave us very valuable lessons in life.
184
638997
3266
10:43
One is this pride in one's own language.
185
643063
3000
์ฒซ์งธ๋กœ ์ž์‹ ์ด ์‚ฌ์šฉํ•˜๋Š” ์–ธ์–ด์— ๋Œ€ํ•œ ์ž๋ถ€์‹ฌ์ž…๋‹ˆ๋‹ค.
10:46
The people of Amale were thrilled to be doing this
186
646063
2667
์•„๋ง๋ฆฌ ์ฃผ๋ฏผ๋“ค์€ ์ €ํฌ ์ผ์„ ์ •๋ง๋กœ ํ•˜๊ณ  ์‹ถ์–ดํ–ˆ๋Š”๋ฐ,
10:48
because they were advancing their own language by doing this.
187
648763
5834
์ด๋ฅผ ํ†ตํ•ด ์ž์‹ ๋“ค์˜ ์–ธ์–ด๋ฅผ ๋ฐœ์ „์‹œํ‚ค๊ณ  ์žˆ์—ˆ๊ธฐ ๋•Œ๋ฌธ์ž…๋‹ˆ๋‹ค.
10:54
The second was the value of community.
188
654630
1833
๋‘ ๋ฒˆ์งธ๋Š” ๊ณต๋™์ฒด์˜ ๊ฐ€์น˜์ž…๋‹ˆ๋‹ค.
10:56
Very quickly, this became a village community effort.
189
656497
3900
์ด ์‚ฌ์—…์€ ์•„์ฃผ ๋น ๋ฅด๊ฒŒ ๋งˆ์„ ์ „์ฒด์˜ ์ผ์ด ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.
11:00
People would gather together in tasks and do this together as a group.
190
660430
4867
์ฃผ๋ฏผ๋“ค์€ ๊ณผ์ œ๋ฅผ ์ค‘์‹ฌ์œผ๋กœ ๋ชจ์ด๊ณ  ์ง‘๋‹จ์œผ๋กœ ๊ณผ์ œ๋ฅผ ์ˆ˜ํ–‰ํ–ˆ์Šต๋‹ˆ๋‹ค.
11:05
And the third is the importance of storytelling.
191
665330
3767
์„ธ ๋ฒˆ์งธ๋Š” ์ด์•ผ๊ธฐํ•˜๊ธฐ์˜ ์ค‘์š”์„ฑ์ž…๋‹ˆ๋‹ค.
11:09
People of Amale were so starved of content that in the morning, during the daytime,
192
669130
6067
์•„๋ง๋ฆฌ ์ฃผ๋ฏผ๋“ค์€ ๋„ˆ๋ฌด๋‚˜๋„ ์ด์•ผ๊นƒ๊ฑฐ๋ฆฌ๊ฐ€ ์—†์—ˆ๊ธฐ์—,
์•„์นจ์ด๋‚˜ ํ•ด๊ฐ€ ๋–  ์žˆ์„ ๋™์•ˆ์—๋Š” ์นด๋ฆฌ์•ผ์— ์ด์•ผ๊ธฐ๋ฅผ ๋…น์Œํ•˜๊ณ 
11:15
they would do recordings of stories in Karya
193
675197
4066
11:19
and then in the evening they would gather the entire village
194
679297
2833
์ €๋…์—๋Š” ์˜จ ๋งˆ์„ ์ „์ฒด๊ฐ€ ๋ชจ์—ฌ์„œ
์ด ์ด์•ผ๊ธฐ๋“ค์„ ๋‹ค์‹œ ๋งํ•˜๊ณ  ๋‹ค์‹œ ์ด์•ผ๊ธฐํ–ˆ์Šต๋‹ˆ๋‹ค.
11:22
and retell and recount these stories to the village.
195
682163
3834
11:27
So as scientists, we get so caught up
196
687563
2167
๊ณผํ•™์ž๋กœ์„œ ์ €ํฌ๋Š” ์ €ํฌ๊ฐ€ ํ•˜๋Š” ์ผ์˜
11:29
in the science and technology part of what we are doing, you know --
197
689763
3700
๊ณผํ•™์ด๋‚˜ ๊ธฐ์ˆ  ๋ฉด์—๋งŒ ์‚ฌ๋กœ์žกํž™๋‹ˆ๋‹ค.
11:33
which is the next best model to have,
198
693497
1966
๋‹ค์Œ ์ตœ๊ณ ์˜ ๋ชจํ˜•์€ ๋ฌด์—‡์ด๊ณ  ์ฒด๊ณ„์˜ ์ •ํ™•๋„๋ฅผ ์–ด๋–ป๊ฒŒ ๋†’์ผ์ง€,
11:35
how can we increase the accuracy of my system,
199
695497
2933
11:38
how can I build the next best system there is --
200
698463
4600
์–ด๋–ป๊ฒŒ ํ•˜๋ฉด ๋‹ค์Œ ์„ธ๋Œ€์˜ ์ตœ๊ณ  ๋ชจํ˜•์„ ๋งŒ๋“ค ์ˆ˜ ์žˆ์„๊นŒ.
11:43
that we forget the reason why we are doing this: the people.
201
703063
3700
๊ทธ๋ž˜์„œ ์™œ ์ด ์ผ์„ ํ•˜๋Š”์ง€ ์žŠ์–ด๋ฒ„๋ฆด ๋•Œ๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค.
๋ฐ”๋กœ ์‚ฌ๋žŒ์ด์ฃ .
11:46
And any successful technology is the one that keeps the people and the users
202
706797
5466
์„ฑ๊ณต์ ์ธ ๊ธฐ์ˆ ์€ ์‚ฌ๋žŒ๋“ค๊ณผ ์‚ฌ์šฉ์ž๋“ค์„
๋‚ด์„ธ์šฐ๊ณ  ๊ด€์‹ฌ๋ฐ›๊ฒŒ ํ•˜๋Š” ๊ธฐ์ˆ ์ž…๋‹ˆ๋‹ค.
11:52
up front and center.
203
712297
1566
11:54
And when they start doing that,
204
714830
1533
๊ธฐ์ˆ ์ด ์ด๋Ÿฐ ์—ญํ• ์„ ํ•˜๊ธฐ ์‹œ์ž‘ํ•˜๋ฉด
11:56
we also realize that technology is probably a very small part of this
205
716363
3667
์šฐ๋ฆฌ๋Š” ๊ธฐ์ˆ ์ด ์•„๋งˆ๋„ ์ด ๊ณผ์ •์—์„œ ์•„์ฃผ ์ž‘์€ ๋ถ€๋ถ„์ž„์„ ๊นจ๋‹ซ๊ณ 
12:00
and there are other things in the story.
206
720030
2367
์ด์•ผ๊ธฐ์— ๋‹ค๋ฅธ ๊ฒƒ๋“ค์ด ์žˆ์Œ์„ ์•Œ๊ฒŒ ๋ฉ๋‹ˆ๋‹ค.
12:02
Maybe there are social, cultural and policy interventions
207
722430
3333
์‚ฌํšŒ์ , ๋ฌธํ™”์ , ์ •์ฑ…์  ๊ฐœ์ž…์ด ๊ธฐ์ˆ ๋งŒํผ ํ•„์š”ํ• ์ง€๋„ ๋ชจ๋ฆ…๋‹ˆ๋‹ค.
12:05
that are required, as much as technology.
208
725797
2200
12:09
So some time back, I worked on a project called VideoKheti
209
729030
3267
์˜ˆ์ „์— ์ €๋Š” ๋น„๋””์˜ค์ผ€ํ‹ฐ๋ผ๋Š” ์‚ฌ์—…์— ์ฐธ์—ฌํ•œ ์ ์ด ์žˆ๋Š”๋ฐ
12:12
that allowed Hindi-speaking farmers in Central India
210
732330
3600
ํžŒ๋””์–ด๋ฅผ ์‚ฌ์šฉํ•˜๋Š” ์ธ๋„ ์ค‘๋ถ€ ๋†๋ถ€๋“ค์ด
12:15
to search for agricultural videos by speaking into a phone-based app.
211
735963
6667
๋†์—… ์˜์ƒ์„ ์Œ์„ฑ์œผ๋กœ ์ฐพ์„ ์ˆ˜ ์žˆ๋Š”
์ „ํ™”๊ธฐ์šฉ ์•ฑ์ด์—ˆ์Šต๋‹ˆ๋‹ค.
12:23
So we went to Madhya Pradesh to collect data for this,
212
743363
3434
์ž๋ฃŒ๋ฅผ ๋ชจ์œผ๊ธฐ ์œ„ํ•ด ๋งˆ๋””์•ผ ํ”„๋ผ๋ฐ์‹œ๋กœ ํ–ฅํ–ˆ์Šต๋‹ˆ๋‹ค.
12:26
and we came back and we were training our models
213
746830
2300
๋Œ์•„์™€์„œ ๋ชจํ˜•๋“ค์„ ํ•™์Šต์‹œํ‚ค๋‹ค๊ฐ€
12:29
and we discovered we're getting very bad results.
214
749163
2367
์•„์ฃผ ์ข‹์ง€ ์•Š์€ ๊ฒฐ๊ณผ๋ฅผ ๋‚ด๊ณ  ์žˆ์Œ์„ ์•Œ์•˜์Šต๋‹ˆ๋‹ค.
12:31
This is not working.
215
751530
1267
์ œ๋Œ€๋กœ ์ž‘๋™ํ•˜์งˆ ์•Š์•˜๊ณ  ์šฐ๋ฆฌ๋Š” ํ—ท๊ฐˆ๋ ธ์Šต๋‹ˆ๋‹ค.
12:32
So we were very confused. Why is this happening?
216
752830
2267
์–ด๋–ป๊ฒŒ ๋œ ์ผ์ด์ง€?
12:35
So we looked deeper and deeper into the data
217
755130
2267
์ž๋ฃŒ๋ฅผ ๋”์šฑ ์„ธ๋ฐ€ํžˆ ๊ด€์ฐฐํ•˜๋‹ค๊ฐ€ ๊นจ๋‹ฌ์•˜์Šต๋‹ˆ๋‹ค.
12:37
and discovered that, yes, we had collected data
218
757430
2233
์šฐ๋ฆฌ๋Š” ๋งˆ์„์ด ์ €๋…์— ์ •๋ง ์กฐ์šฉํ•˜๊ณ  ๊ณ ์š”ํ•  ๊ฒƒ์œผ๋กœ ์ƒ๊ฐํ–ˆ์Šต๋‹ˆ๋‹ค.
12:39
from what we thought was a very silent, quiet village in the evening.
219
759697
4666
12:44
But what we hadn't heard while we were doing this
220
764363
3400
๊ทธ๋Ÿฌ๋‚˜ ์ด ์ผ์„ ์ง„ํ–‰ํ•˜๋ฉด์„œ ์ €ํฌ๊ฐ€ ๋ชฐ๋ž๋˜ ๊ฒƒ์€
12:47
was that there was this constant buzz of night insects, you know?
221
767797
4100
์•ผํ–‰์„ฑ ๊ณค์ถฉ๋“ค์˜ ์ง€์ง€์ง ์†Œ๋ฆฌ๊ฐ€ ๋Š์ด์ง€ ์•Š์•˜๋‹ค๋Š” ๊ฒƒ์ด์ฃ .
12:51
So throughout the recordings, we had this "bzz" of the insects,
222
771930
3467
๊ทธ๋ž˜์„œ ๋…น์Œ ๊ณณ๊ณณ์— ๊ณค์ถฉ ์†Œ๋ฆฌ๊ฐ€ ๋“ค์–ด ์žˆ์—ˆ๊ณ ,
12:55
which was actually distorting our speech.
223
775430
2533
์ด ์†Œ๋ฆฌ๊ฐ€ ์Œ์„ฑ์„ ์™œ๊ณกํ•˜๊ณ  ์žˆ์—ˆ์Šต๋‹ˆ๋‹ค.
12:58
The second thing was that when we went there
224
778797
2233
๋‘ ๋ฒˆ์งธ๋Š” ์ €ํฌ๊ฐ€ ๋งˆ์„๋กœ ๊ฐ€์„œ
13:01
to kind of test our app in the village,
225
781030
3867
๋งˆ์„ ๋‚ด์—์„œ ์•ฑ์„ ์‹œํ—˜ํ•˜๋ ค๊ณ  ํ–ˆ์„ ๋•Œ,
13:04
I and my colleague Indrani Medhi,
226
784930
2533
์ œ ๋™๋ฃŒ์ด์ž ๋Šฅ๋ ฅ์„ ์ธ์ •๋ฐ›๋Š” ๋„์•ˆ ์—ฐ๊ตฌ์ž์ธ ์ธ๋“œ๋ผ๋‹ˆ ๋ฉ”๋””์™€ ์ €๋Š”
13:07
who is a very well-regarded design researcher,
227
787497
3600
13:11
we found that the women couldn't pronounce the sanskritized words
228
791130
4400
์—ฌ์„ฑ๋“ค์ด ๊ฒ€์ƒ‰์–ด ์ค‘ ์‚ฐ์Šคํฌ๋ฆฌํŠธ์–ด๋กœ ๋ฒˆ์—ญ๋œ ๋‹จ์–ด๋“ค์„
์ฝ์ง€ ๋ชปํ•œ๋‹ค๋Š” ๊ฒƒ์„ ๋ฐœ๊ฒฌํ–ˆ์Šต๋‹ˆ๋‹ค.
13:15
that we had for some of the search terms.
229
795530
2767
13:18
So, like ...
230
798330
1500
๊ทธ๋Ÿฌ๋‹ˆ๊นŒโ€ฆ
13:21
(speaks Hindi)
231
801663
2800
(ํžŒ๋””์–ด)
13:24
Which is like the term for chemical pesticides, right?
232
804497
3866
์ด๊ฒƒ์€ ํžŒ๋””์–ด๋กœ ํ™”ํ•™ ์‚ด์ถฉ์ œ์ž…๋‹ˆ๋‹ค.
13:28
Because we got these terms from the agricultural extension center
233
808363
5534
๋†์—… ์ง„ํฅ ์„ผํ„ฐ์—์„œ ์ด ์šฉ์–ด๋“ค์„ ์ˆ˜์ง‘ํ–ˆ๋Š”๋ฐ
13:33
and the women, even though they are farming,
234
813930
2100
์—ฌ์„ฑ๋“ค์€ ๋†์—…์— ์ข…์‚ฌํ•˜๋”๋ผ๋„ ์„ผํ„ฐ์™€์˜ ์ ‘์ด‰์ด ์ „ํ˜€ ์—†์Šต๋‹ˆ๋‹ค.
13:36
do not interact with that center at all.
235
816030
2867
13:38
The men do, the women probably use something much simpler, like ...
236
818930
3967
๋‚จ์„ฑ๋“ค์ด ์ ‘์ด‰ํ•ฉ๋‹ˆ๋‹ค.
์—ฌ์„ฑ๋“ค์€ ํ›จ์”ฌ ๊ฐ„๋‹จํ•œ ์šฉ์–ด๋ฅผ ์‚ฌ์šฉํ•  ๊ฒƒ์ž…๋‹ˆ๋‹ค.
13:42
(speaks Hindi)
237
822930
1300
(ํžŒ๋””์–ด)
13:44
Which basically means killing pests with medicine.
238
824263
3534
๊ฐ„๋‹จํžˆ ๋งํ•˜๋ฉด ํ•ด์ถฉ์„ ์•ฝ์œผ๋กœ ์ฃฝ์ธ๋‹ค๋Š” ๋œป์ž…๋‹ˆ๋‹ค.
13:48
So what I have learned through my journey
239
828430
3867
์ œ๊ฐ€ ์ด ์—ฌํ–‰์„ ํ†ตํ•ด ๋ฐฐ์› ๊ณ 
13:52
and what I would like to put across to you --
240
832330
2900
์—ฌ๋Ÿฌ๋ถ„๊ป˜ ์ „ํ•ด๋“œ๋ฆฌ๊ณ  ์‹ถ์€ ๊ฒƒ์€
13:55
by now, I hope you've understood me,
241
835263
2000
์ง€๊ธˆ์ฏค์€ ์ œ ๋ง์ด ์ดํ•ด๊ฐ€ ๋˜์…จ์œผ๋ฉด ์ข‹๊ฒ ๋Š”๋ฐ,
13:57
is that there is the majority of the world's languages
242
837297
3433
์„ธ๊ณ„ ์–ธ์–ด ์ค‘ ๋Œ€๋ถ€๋ถ„์€
14:00
that require intensive investment for resource creation
243
840763
4267
์–ธ์–ด ๊ธฐ์ˆ ๋กœ ์ด์ต์„ ์–ป๊ธฐ ์œ„ํ•ด์„œ๋Š” ์ž์› ์ƒ์„ฑ์— ์ง‘์ค‘ ํˆฌ์ž๊ฐ€ ํ•„์š”ํ•ฉ๋‹ˆ๋‹ค.
14:05
if they are to benefit from language technology.
244
845030
2567
14:07
And this is unlikely to happen in a very fast and efficient manner.
245
847630
5367
์ด๋Š” ์•„์ฃผ ๋น ๋ฅด๊ณ  ํšจ์œจ์ ์œผ๋กœ ์ด๋ฃจ์–ด์งˆ ๊ฒƒ ๊ฐ™์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
14:13
So it is extremely important for us to ensure
246
853963
2934
๊ทธ๋Ÿฌ๋ฏ€๋กœ ์–ธ์–ด ๊ธฐ์ˆ  ๋ถ„์•ผ์—์„œ ์ €ํฌ๊ฐ€ ํ•˜๋Š” ์ผ์„ ํ†ตํ•ด
14:16
that the community derives maximum benefit
247
856930
3533
์‚ฌํšŒ๊ฐ€ ์ตœ๋Œ€์˜ ์ด์ต์„ ์–ป๋„๋ก ํ•˜๋Š” ์ผ์€
14:20
from whatever that we are doing in the language tech area.
248
860497
3966
๋งค์šฐ ์ค‘์š”ํ•œ ์ผ์ž…๋‹ˆ๋‹ค.
14:24
And to do this and deliver a positive social impact
249
864497
3466
์ด๋ฅผ ๋ณด์žฅํ•˜๊ณ  ํ•ด๋‹น ์‚ฌํšŒ์— ๊ธ์ •์  ์˜ํ–ฅ์„ ์ฃผ๊ธฐ ์œ„ํ•ด์„œ
14:27
on these communities,
250
867997
1466
14:29
we follow what we call the modified 4-D design thinking methodology.
251
869497
4733
์ˆ˜์ •ํ•œ 4-D ๋””์ž์ธ ์‚ฌ๊ณ  ๋ฐฉ๋ฒ•๋ก ์ด๋ผ ๋ถ€๋ฅด๋Š” ๋ฐฉ๋ฒ•์„ ์‚ฌ์šฉํ•ฉ๋‹ˆ๋‹ค.
14:34
So the 4-D means: discover, design, develop and deploy.
252
874263
5200
๋„ค๊ฐ€์ง€์˜ D๋Š” ๋ฐœ๊ฒฌ(discover),
๋””์ž์ธ(design), ๋ฐœ์ „(develop), ๊ทธ๋ฆฌ๊ณ  ์ ์šฉ(deploy)์ž…๋‹ˆ๋‹ค.
14:39
So discover the problem that language technology can solve
253
879497
3066
ํŠน์ • ์–ธ์–ด ๊ณต๋™์ฒด์—์„œ ์–ธ์–ด ๊ธฐ์ˆ ์ด ํ•ด๊ฒฐ ๊ฐ€๋Šฅํ•œ ๋ฌธ์ œ๋ฅผ ๋ฐœ๊ฒฌํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
14:42
for a particular language community.
254
882563
2200
14:44
This observation-led approach can help allocate resources
255
884797
3233
๊ด€์ฐฐ ์ค‘์‹ฌ์  ์ ‘๊ทผ์€ ์ œ์ผ ํ•„์š”ํ•œ ๊ณณ์— ์ž์›์„ ๋ฐฐ๋ถ„ํ•˜๋Š” ๊ฑธ ๋„์šธ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
14:48
where they are most needed,
256
888030
1700
14:49
designed for the users and their language,
257
889763
2767
์–ธ์–ด์™€ ์–ธ์–ด ์‚ฌ์šฉ์ž๋ฅผ ์œ„ํ•ด ๊ณ ์•ˆ๋˜์–ด์„œ
14:52
understand the diversity in the linguistic properties
258
892530
3367
์–ธ์–ด ์š”์†Œ์˜ ๋‹ค์–‘์„ฑ๊ณผ ์„ธ๊ณ„ ์–ธ์–ด์˜ ๋‹ค์–‘์„ฑ์„ ์ดํ•ดํ•˜๋Š” ๋ฐ ๋„์›€์„ ์ค๋‹ˆ๋‹ค.
14:55
and the languages of the world.
259
895930
2100
14:58
And don't think, oh, this is made for English.
260
898030
2400
โ€˜์ด๊ฑด ์˜์–ด๋ฅผ ์œ„ํ•ด ๋งŒ๋“ค์—ˆ๋Š”๋ฐ ๋งˆ๋ผํ‹ฐ์–ด๋‚˜ ๊ณค๋“œ์–ด์— ์–ด๋–ป๊ฒŒ ์ ์šฉํ•˜์ง€?โ€™
15:00
Now, how can we just adapt it for Marathi or for Gondi, right?
261
900463
4334
๋ผ๊ณ  ์ƒ๊ฐํ•˜์ง€ ๋ง์•„์ฃผ์„ธ์š”.
15:04
Develop rapidly and deploy frequently.
262
904830
2600
๋น ๋ฅด๊ฒŒ ๊ฐœ๋ฐœํ•˜๊ณ  ์ž์ฃผ ์ ์šฉํ•ฉ๋‹ˆ๋‹ค.
15:07
It's an iterative process that will help you fail fast
263
907463
3500
๋ฐ˜๋ณต์ ์ธ ๊ณผ์ •์œผ๋กœ์„œ ์‚ฌ์šฉ์ž๊ฐ€ ๋น ๋ฅด๊ฒŒ ์‹คํŒจํ•˜๋„๋ก ๋•๊ณ 
15:10
and early failures will eventually lead to success.
264
910997
3366
์ด๋ฅธ ์‹คํŒจ๋Š” ๊ฒฐ๊ตญ ์„ฑ๊ณต์ด ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค.
15:15
The important thing is to persevere.
265
915497
1966
์ค‘์š”ํ•œ ์ ์€ ์ธ๋‚ดํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
15:17
Do not give up.
266
917497
1366
ํฌ๊ธฐํ•˜์ง€ ๋งˆ์‹ญ์‹œ์˜ค.
15:18
And I remember the story of these two Aborigine Australian women,
267
918863
5734
์˜ค์ŠคํŠธ๋ ˆ์ผ๋ฆฌ์•„ ์›์ฃผ๋ฏผ ์—ฌ์„ฑ ๋‘ ๋ช…์˜ ์ด์•ผ๊ธฐ๊ฐ€ ๊ธฐ์–ต๋‚ฉ๋‹ˆ๋‹ค.
15:24
Patricia O'Connor and Ysola Best.
268
924630
3800
ํŽ˜ํŠธ๋ฆฌ์ƒค ์˜ค์ฝ”๋„ˆ์™€ ์ด์†”๋ผ ๋ฒ ์ŠคํŠธ์˜ ์ด์•ผ๊ธฐ์ž…๋‹ˆ๋‹ค.
15:29
In the mid-90s, they went to the University of Queensland
269
929763
3134
1990๋…„๋Œ€ ์ค‘๋ฐ˜ ๊ทธ๋“ค์€ ํ€ธ์ฆ๋žœ๋“œ ๋Œ€ํ•™์— ๊ฐ”๊ณ 
15:32
and they wanted to learn their own language, called Yugambeh,
270
932930
3333
๊ทธ๋“ค์˜ ์–ธ์–ด์ธ ์œ ๊ฐ๋ฒ ์–ด๋ฅผ ๋ฐฐ์šฐ๊ณ  ์‹ถ์—ˆ์Šต๋‹ˆ๋‹ค.
15:36
and they were told very bluntly, "Your language is dead.
271
936297
2633
์•„์ฃผ ํ‰๋ช…์Šค๋Ÿฝ๊ฒŒ ์ด๋Ÿฐ ๋‹ต์ด ๋Œ์•„์™”์Šต๋‹ˆ๋‹ค.
15:38
It's been dead for three decades.
272
938963
1600
โ€œ์—ฌ๋Ÿฌ๋ถ„์˜ ์–ธ์–ด๋Š” ์—†์–ด์กŒ๊ณ  ์‚ฌ์–ด๊ฐ€ ๋œ ์ง€ 30๋…„์ด ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.
15:40
You cannot work on this. Find something else to work on."
273
940563
3867
์ด ๋ถ„์•ผ๋ฅผ ๊ณต๋ถ€ํ•  ์ˆ˜ ์—†์œผ๋‹ˆ ๋‹ค๋ฅธ ๋ถ„์•ผ๋ฅผ ์ฐพ์•„๋ณด์„ธ์š”.โ€
15:44
They did not give up.
274
944463
1267
๊ทธ๋“ค์€ ํฌ๊ธฐํ•˜์ง€ ์•Š์•˜์Šต๋‹ˆ๋‹ค.
15:45
They went to the community,
275
945763
1600
๊ทธ๋“ค์€ ์œ ๊ฐ๋ฒ ์–ด ๊ณต๋™์ฒด๋ฅผ ๋ฐฉ๋ฌธํ–ˆ๊ณ 
15:47
they dug up oral memories, oral traditions, oral literature,
276
947363
4867
๋ง๋กœ ์ „ํ•ด์ง€๋Š” ๊ธฐ์–ต, ์ „ํ†ต๊ณผ ๋ฌธํ•™์„ ๋ฐœ๊ตดํ–ˆ์œผ๋ฉฐ
15:52
and founded the Yugambeh Museum,
277
952263
3367
์œ ๊ฐ๋ฒ  ๋ฐ•๋ฌผ๊ด€์„ ์„ค๋ฆฝํ–ˆ์Šต๋‹ˆ๋‹ค.
15:55
which became the most important cultural and linguistic center for the language
278
955663
5434
์ด ๋ฐ•๋ฌผ๊ด€์€ ์œ ๊ฐ๋ฒ ์–ด์™€ ๊ทธ ๊ณต๋™์ฒด์—๊ฒŒ ๋ฌธํ™”์ ์œผ๋กœ
๊ทธ๋ฆฌ๊ณ  ์–ธ์–ด์ ์œผ๋กœ ๊ฐ€์žฅ ์ค‘์š”ํ•œ ๊ณณ์ด ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.
16:01
and its community.
279
961130
1767
16:02
They did not have technology. They only had their willpower.
280
962930
4033
๊ทธ๋“ค์€ ๊ธฐ์ˆ ์ด ์—†์—ˆ์Šต๋‹ˆ๋‹ค.
์˜ค์ง ์˜์ง€๋งŒ ์žˆ์—ˆ์ฃ .
16:06
Now, with the power of technology,
281
966997
2233
์ด์ œ ๊ธฐ์ˆ ์˜ ํž˜์œผ๋กœ
16:09
we can ensure that the next page is written in Salmi from Finland,
282
969263
5767
์šฐ๋ฆฌ๋Š” ์ด๋“ค ์–ธ์–ด์˜ ๋ฏธ๋ž˜๋ฅผ ๋ณด์žฅํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
ํ•€๋ž€๋“œ์˜ ์‚ด๋ฏธ์–ด,
16:15
Lillooet from Canada or Mundari from India.
283
975030
3467
์บ๋‚˜๋‹ค์˜ ๋ฆด๋ฃจ์—ฃ์–ด, ์ธ๋„์˜ ๋ฌธ๋‹ค๋ฆฌ์–ด.
16:19
Thank you.
284
979163
1000
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค.
์ด ์›น์‚ฌ์ดํŠธ ์ •๋ณด

์ด ์‚ฌ์ดํŠธ๋Š” ์˜์–ด ํ•™์Šต์— ์œ ์šฉํ•œ YouTube ๋™์˜์ƒ์„ ์†Œ๊ฐœํ•ฉ๋‹ˆ๋‹ค. ์ „ ์„ธ๊ณ„ ์ตœ๊ณ ์˜ ์„ ์ƒ๋‹˜๋“ค์ด ๊ฐ€๋ฅด์น˜๋Š” ์˜์–ด ์ˆ˜์—…์„ ๋ณด๊ฒŒ ๋  ๊ฒƒ์ž…๋‹ˆ๋‹ค. ๊ฐ ๋™์˜์ƒ ํŽ˜์ด์ง€์— ํ‘œ์‹œ๋˜๋Š” ์˜์–ด ์ž๋ง‰์„ ๋”๋ธ” ํด๋ฆญํ•˜๋ฉด ๊ทธ๊ณณ์—์„œ ๋™์˜์ƒ์ด ์žฌ์ƒ๋ฉ๋‹ˆ๋‹ค. ๋น„๋””์˜ค ์žฌ์ƒ์— ๋งž์ถฐ ์ž๋ง‰์ด ์Šคํฌ๋กค๋ฉ๋‹ˆ๋‹ค. ์˜๊ฒฌ์ด๋‚˜ ์š”์ฒญ์ด ์žˆ๋Š” ๊ฒฝ์šฐ ์ด ๋ฌธ์˜ ์–‘์‹์„ ์‚ฌ์šฉํ•˜์—ฌ ๋ฌธ์˜ํ•˜์‹ญ์‹œ์˜ค.

https://forms.gle/WvT1wiN1qDtmnspy7