1
00:00:00,480 --> 00:00:03,480
foreign

2
00:00:08,720 --> 00:00:15,780
to the our connected Universe track

3
00:00:12,139 --> 00:00:18,779
AKA everything about apis we're so

4
00:00:15,780 --> 00:00:20,100
stoked that uh I can I can Glimpse at so

5
00:00:18,779 --> 00:00:22,980
many people's faces through the light

6
00:00:20,100 --> 00:00:25,740
and we're really looking forward to uh

7
00:00:22,980 --> 00:00:26,699
all of the talks in the truck today uh

8
00:00:25,740 --> 00:00:28,140
now we're running a little bit behind

9
00:00:26,699 --> 00:00:29,939
schedule after the opening address so

10
00:00:28,140 --> 00:00:32,700
we're going to kick right into our first

11
00:00:29,939 --> 00:00:36,420
talk of the day and we have up our first

12
00:00:32,700 --> 00:00:38,660
Anushka vidanade uh sorry and just just

13
00:00:36,420 --> 00:00:40,980
Moore and they're going to be presenting

14
00:00:38,660 --> 00:00:43,140
building Health software on

15
00:00:40,980 --> 00:00:45,719
decentralized health data using personal

16
00:00:43,140 --> 00:00:49,260
online data stores I'm really excited

17
00:00:45,719 --> 00:00:53,399
for this talk take it away thanks yeah

18
00:00:49,260 --> 00:00:56,219
right uh good morning everyone so yeah

19
00:00:53,399 --> 00:00:58,320
first um I would like to begin today by

20
00:00:56,219 --> 00:01:01,079
acknowledging the traditional custodians

21
00:00:58,320 --> 00:01:02,579
of the land on which we meet today and

22
00:01:01,079 --> 00:01:06,360
would like to pay my respect to their

23
00:01:02,579 --> 00:01:08,040
Ides past present and emerging and we

24
00:01:06,360 --> 00:01:10,080
would also like to acknowledge our

25
00:01:08,040 --> 00:01:14,760
indigenous Partners on this project that

26
00:01:10,080 --> 00:01:18,060
we are doing at Anu why who these are

27
00:01:14,760 --> 00:01:20,280
from yaraba no not Queensland

28
00:01:18,060 --> 00:01:24,439
the traditional owners gugganji people

29
00:01:20,280 --> 00:01:24,439
and the history corners of that land

30
00:01:24,860 --> 00:01:30,600
and here with me is just Moe and we are

31
00:01:28,680 --> 00:01:33,119
from the softer Innovation Institute

32
00:01:30,600 --> 00:01:35,340
School of computing from the Australian

33
00:01:33,119 --> 00:01:38,159
National University and today we'll be

34
00:01:35,340 --> 00:01:39,659
talking about personal online data

35
00:01:38,159 --> 00:01:43,380
stores and how do we build Health

36
00:01:39,659 --> 00:01:45,560
software using personal online data

37
00:01:43,380 --> 00:01:45,560
stores

38
00:01:46,640 --> 00:01:54,420
and this is a brief outline of our talk

39
00:01:50,220 --> 00:01:55,740
so I'm gonna do whys and what's of these

40
00:01:54,420 --> 00:01:57,840
uh

41
00:01:55,740 --> 00:02:01,140
something called pods personal online

42
00:01:57,840 --> 00:02:03,960
data stores and solid boards and I'm

43
00:02:01,140 --> 00:02:05,939
gonna dive a little bit deep uh into

44
00:02:03,960 --> 00:02:09,300
some technologies Behind these solid

45
00:02:05,939 --> 00:02:12,060
pods and then Jess will talk about our

46
00:02:09,300 --> 00:02:14,819
actual project Where We Are

47
00:02:12,060 --> 00:02:17,300
build this health software using solid

48
00:02:14,819 --> 00:02:17,300
pods

49
00:02:17,819 --> 00:02:22,920
okay

50
00:02:18,959 --> 00:02:26,819
so uh not sure if everyone can see that

51
00:02:22,920 --> 00:02:28,500
so what I did was I asked chat GPT what

52
00:02:26,819 --> 00:02:32,520
are the problems with centralized

53
00:02:28,500 --> 00:02:35,459
databases and here's what I got

54
00:02:32,520 --> 00:02:38,640
um so there's a long list but I want you

55
00:02:35,459 --> 00:02:41,459
to focus on these three aspect so

56
00:02:38,640 --> 00:02:43,980
basically these three aspect highlights

57
00:02:41,459 --> 00:02:47,700
uh the aspect of data privacy and

58
00:02:43,980 --> 00:02:50,879
security and lack of transparency

59
00:02:47,700 --> 00:02:54,360
and uh the lack of data ownership

60
00:02:50,879 --> 00:02:56,160
so the Privacy as evident by latest data

61
00:02:54,360 --> 00:02:58,260
breaches all over the world here in

62
00:02:56,160 --> 00:03:01,140
Australia as well for instance Medicare

63
00:02:58,260 --> 00:03:03,000
and Optus as some of you may know so the

64
00:03:01,140 --> 00:03:06,300
centralized databases are prone to

65
00:03:03,000 --> 00:03:08,760
security and privacy risks always so one

66
00:03:06,300 --> 00:03:12,060
of the reasons being that they can be

67
00:03:08,760 --> 00:03:14,940
these centralized databases because they

68
00:03:12,060 --> 00:03:16,980
hold a lot of data about uh let's say

69
00:03:14,940 --> 00:03:18,959
hundreds of thousands of or even

70
00:03:16,980 --> 00:03:22,260
millions of people they can be

71
00:03:18,959 --> 00:03:24,480
attractive many adversaries out there

72
00:03:22,260 --> 00:03:26,940
um so a successful bridge to these

73
00:03:24,480 --> 00:03:30,000
centralized databases could reveal

74
00:03:26,940 --> 00:03:30,780
sensitive information about yourself

75
00:03:30,000 --> 00:03:33,000
um

76
00:03:30,780 --> 00:03:36,060
so and obviously there's a there's a

77
00:03:33,000 --> 00:03:38,640
threat of privacy where our own data is

78
00:03:36,060 --> 00:03:41,340
most of the time handled by someone else

79
00:03:38,640 --> 00:03:44,280
some Central Authority

80
00:03:41,340 --> 00:03:47,819
right and then the lack of transparency

81
00:03:44,280 --> 00:03:51,120
centralized databases often rely on

82
00:03:47,819 --> 00:03:54,060
um Central authorities or administrators

83
00:03:51,120 --> 00:03:57,000
which we call them and we basically put

84
00:03:54,060 --> 00:03:59,760
our data on their hands and we basically

85
00:03:57,000 --> 00:04:02,459
trust them uh because sometimes that

86
00:03:59,760 --> 00:04:05,400
could be our only option right and in

87
00:04:02,459 --> 00:04:08,280
return we do not really get to know what

88
00:04:05,400 --> 00:04:10,680
happens with our data and what kind of

89
00:04:08,280 --> 00:04:14,519
raw data really they capture about

90
00:04:10,680 --> 00:04:16,919
ourselves and what do they do with them

91
00:04:14,519 --> 00:04:19,919
so even though the Privacy policies such

92
00:04:16,919 --> 00:04:22,500
as gdpr enforce the practice of

93
00:04:19,919 --> 00:04:25,979
transparency there will still be gaps

94
00:04:22,500 --> 00:04:28,800
with these centralized databases

95
00:04:25,979 --> 00:04:31,800
um you might remember the the incidents

96
00:04:28,800 --> 00:04:34,680
with Cambridge analytical Scandal so

97
00:04:31,800 --> 00:04:37,800
yeah so there's lack of transparency and

98
00:04:34,680 --> 00:04:40,020
finally the lack of ownership uh

99
00:04:37,800 --> 00:04:44,040
so we when relying on these third

100
00:04:40,020 --> 00:04:46,680
parties to handle our own private data

101
00:04:44,040 --> 00:04:49,620
um we have to we don't really get the

102
00:04:46,680 --> 00:04:52,440
ownership of our own data right what we

103
00:04:49,620 --> 00:04:55,320
get most of the time with these systems

104
00:04:52,440 --> 00:04:58,020
is some kind of a statement or a

105
00:04:55,320 --> 00:05:00,300
document stating that they handle they

106
00:04:58,020 --> 00:05:02,520
gather this kind of data of ourselves

107
00:05:00,300 --> 00:05:05,880
and this is what they are doing we don't

108
00:05:02,520 --> 00:05:07,500
actually get to see the raw data itself

109
00:05:05,880 --> 00:05:09,180
yeah so

110
00:05:07,500 --> 00:05:10,800
and that comes with the lack of

111
00:05:09,180 --> 00:05:13,320
interoperability with centralized

112
00:05:10,800 --> 00:05:15,780
databases uh we are dependent on one

113
00:05:13,320 --> 00:05:17,880
database for one service where we cannot

114
00:05:15,780 --> 00:05:19,199
really migrate between databases if we

115
00:05:17,880 --> 00:05:20,160
want to

116
00:05:19,199 --> 00:05:22,320
um

117
00:05:20,160 --> 00:05:25,500
so these are the these are some of the

118
00:05:22,320 --> 00:05:29,280
main issues with centralized databases

119
00:05:25,500 --> 00:05:31,080
so can we somehow solve these issues so

120
00:05:29,280 --> 00:05:32,400
address these issues can we somehow turn

121
00:05:31,080 --> 00:05:36,300
this around

122
00:05:32,400 --> 00:05:38,820
and that's where this the concept of

123
00:05:36,300 --> 00:05:41,280
um puts personal online data stores

124
00:05:38,820 --> 00:05:44,639
comes into play uh

125
00:05:41,280 --> 00:05:46,919
so a personal online data so data store

126
00:05:44,639 --> 00:05:50,220
is a place where you can store your own

127
00:05:46,919 --> 00:05:52,800
personal data and also have complete

128
00:05:50,220 --> 00:05:53,460
control over that data

129
00:05:52,800 --> 00:05:56,460
um

130
00:05:53,460 --> 00:05:57,900
so the concept of PODS is framed around

131
00:05:56,460 --> 00:06:03,080
the individual

132
00:05:57,900 --> 00:06:07,199
so the individual is at the center

133
00:06:03,080 --> 00:06:09,660
as you can see here in this picture and

134
00:06:07,199 --> 00:06:12,240
then individual can access their own

135
00:06:09,660 --> 00:06:14,580
data in their own code they can see

136
00:06:12,240 --> 00:06:17,060
what's been stored they are the one who

137
00:06:14,580 --> 00:06:19,560
store this data and then

138
00:06:17,060 --> 00:06:22,340
individuals are better informed about

139
00:06:19,560 --> 00:06:26,039
their data and

140
00:06:22,340 --> 00:06:28,860
the service providers they come at

141
00:06:26,039 --> 00:06:31,440
around the individual here

142
00:06:28,860 --> 00:06:33,780
so they provide different Services based

143
00:06:31,440 --> 00:06:37,080
on the data that's been stored inside

144
00:06:33,780 --> 00:06:40,020
our personal online data store so that's

145
00:06:37,080 --> 00:06:43,080
the basic concept of uh

146
00:06:40,020 --> 00:06:46,440
you know the personal online data stores

147
00:06:43,080 --> 00:06:49,620
so it's just not only a data store as

148
00:06:46,440 --> 00:06:51,800
well so it's an ecosystem where programs

149
00:06:49,620 --> 00:06:55,440
programmers developers like you and me

150
00:06:51,800 --> 00:06:59,220
can quickly build stuff programs

151
00:06:55,440 --> 00:07:02,639
softwares applications and get our

152
00:06:59,220 --> 00:07:05,100
innovative ideas out in the field so we

153
00:07:02,639 --> 00:07:08,520
can build software to use the data on

154
00:07:05,100 --> 00:07:11,400
pods as a third-party application

155
00:07:08,520 --> 00:07:13,259
so that's the basic idea

156
00:07:11,400 --> 00:07:17,280
um

157
00:07:13,259 --> 00:07:20,940
right so what is solid bones so solid is

158
00:07:17,280 --> 00:07:24,199
basically uh it's uh so it's it's a

159
00:07:20,940 --> 00:07:26,759
short for social link data uh it's a web

160
00:07:24,199 --> 00:07:28,979
decentralization project led by Tim

161
00:07:26,759 --> 00:07:30,960
berners Lee it's initially released in

162
00:07:28,979 --> 00:07:34,020
2016.

163
00:07:30,960 --> 00:07:36,660
um and it contains a specification that

164
00:07:34,020 --> 00:07:38,580
lets individuals store their data

165
00:07:36,660 --> 00:07:40,800
securely in a decentralized databases

166
00:07:38,580 --> 00:07:43,500
these spots

167
00:07:40,800 --> 00:07:46,919
um so it's a place where you can have

168
00:07:43,500 --> 00:07:50,039
your own data stored in your own port

169
00:07:46,919 --> 00:07:53,460
and starts uh you know start

170
00:07:50,039 --> 00:07:55,819
um doing stuff with those data so start

171
00:07:53,460 --> 00:07:59,460
giving access to others on your data

172
00:07:55,819 --> 00:08:01,620
things like that so following solid

173
00:07:59,460 --> 00:08:04,680
specifications multiple servers has been

174
00:08:01,620 --> 00:08:06,419
built so multiple open source solid

175
00:08:04,680 --> 00:08:09,360
servers are available out there so the

176
00:08:06,419 --> 00:08:12,360
most popular one is the community solid

177
00:08:09,360 --> 00:08:15,240
server which is here in the first link

178
00:08:12,360 --> 00:08:17,940
and then you have PHP solid server or

179
00:08:15,240 --> 00:08:19,440
you also have solid server as a plugin

180
00:08:17,940 --> 00:08:22,139
for next cloud

181
00:08:19,440 --> 00:08:25,379
so the difference with all these servers

182
00:08:22,139 --> 00:08:26,400
the concept of the solid servers is that

183
00:08:25,379 --> 00:08:29,280
uh

184
00:08:26,400 --> 00:08:31,560
for the solid specification compared to

185
00:08:29,280 --> 00:08:33,779
centralized options ah centralized

186
00:08:31,560 --> 00:08:36,419
options is that the centralized options

187
00:08:33,779 --> 00:08:38,580
is always

188
00:08:36,419 --> 00:08:41,279
um get a separate backend

189
00:08:38,580 --> 00:08:43,740
right for separate services so you have

190
00:08:41,279 --> 00:08:46,740
a separate backend for separate services

191
00:08:43,740 --> 00:08:50,339
for instance uh Facebook has their own

192
00:08:46,740 --> 00:08:53,940
backend Twitter has their own backend

193
00:08:50,339 --> 00:08:57,060
um but with the solid pods what happens

194
00:08:53,940 --> 00:08:59,700
is that you will have a single back end

195
00:08:57,060 --> 00:09:02,160
which is your own port and the apps and

196
00:08:59,700 --> 00:09:05,940
the services will be asking for your

197
00:09:02,160 --> 00:09:09,959
permission to you use your own data uh

198
00:09:05,940 --> 00:09:12,380
and in return give the service to you

199
00:09:09,959 --> 00:09:12,380
right

200
00:09:12,420 --> 00:09:18,560
so I'm gonna give a brief introduction

201
00:09:15,240 --> 00:09:20,940
to technology behind solid uh solid pods

202
00:09:18,560 --> 00:09:22,380
and then hand over to Jess to talk about

203
00:09:20,940 --> 00:09:25,920
our actual project

204
00:09:22,380 --> 00:09:28,500
so the technology behind pod pods solid

205
00:09:25,920 --> 00:09:31,500
pods is mostly based on semantic web

206
00:09:28,500 --> 00:09:33,600
Technologies so for each Port we do have

207
00:09:31,500 --> 00:09:38,459
something called a web ID

208
00:09:33,600 --> 00:09:41,940
uh which is essentially an URI which is

209
00:09:38,459 --> 00:09:46,019
unique to each pod so each pod can be

210
00:09:41,940 --> 00:09:48,420
uniquely identified using this web ID

211
00:09:46,019 --> 00:09:51,240
um so for instance here you can see this

212
00:09:48,420 --> 00:09:55,800
is my so the first link here that's my

213
00:09:51,240 --> 00:09:57,959
web ID from my pod in in wrap.com so

214
00:09:55,800 --> 00:10:00,300
that's an uh free service which you can

215
00:09:57,959 --> 00:10:02,760
go there and build your own port and

216
00:10:00,300 --> 00:10:05,160
start storing your data so that's the

217
00:10:02,760 --> 00:10:08,640
web ID for my pod

218
00:10:05,160 --> 00:10:12,360
and the pods so the data inputs are

219
00:10:08,640 --> 00:10:15,300
saved as something called link data link

220
00:10:12,360 --> 00:10:17,220
data is basically a collection of terms

221
00:10:15,300 --> 00:10:19,680
um or their definitions and their links

222
00:10:17,220 --> 00:10:21,000
the relationships between them so for

223
00:10:19,680 --> 00:10:24,720
instance here

224
00:10:21,000 --> 00:10:27,420
you can see uh the link graph

225
00:10:24,720 --> 00:10:29,880
so that that's a link to graph uh where

226
00:10:27,420 --> 00:10:34,080
I do you can see relationships between

227
00:10:29,880 --> 00:10:37,500
Bob and Alice and to some of these terms

228
00:10:34,080 --> 00:10:40,500
here so these linked graphs once we have

229
00:10:37,500 --> 00:10:43,320
this linked graph we can store this in

230
00:10:40,500 --> 00:10:45,899
our ports using something called Turtle

231
00:10:43,320 --> 00:10:50,339
serialization so what happens is in

232
00:10:45,899 --> 00:10:51,060
total serialization what we do is we

233
00:10:50,339 --> 00:10:53,700
um

234
00:10:51,060 --> 00:10:55,860
we convert this link graph into

235
00:10:53,700 --> 00:10:58,320
something called triplets we are in

236
00:10:55,860 --> 00:11:00,800
triplets we do have the subjects

237
00:10:58,320 --> 00:11:03,899
predicates and object so this link graph

238
00:11:00,800 --> 00:11:06,660
is now converted into these triplets

239
00:11:03,899 --> 00:11:09,360
like the ones you see there and then

240
00:11:06,660 --> 00:11:11,220
that's been stored in Turtle files and

241
00:11:09,360 --> 00:11:13,800
there's those Turtle files are inside

242
00:11:11,220 --> 00:11:16,860
your ports and that's your data

243
00:11:13,800 --> 00:11:19,620
so that's the basic idea of storing data

244
00:11:16,860 --> 00:11:21,680
in solid pods

245
00:11:19,620 --> 00:11:21,680
um

246
00:11:22,140 --> 00:11:26,700
and then

247
00:11:23,880 --> 00:11:30,000
um so the authentication stuff and the

248
00:11:26,700 --> 00:11:31,980
authorization stops so in solid ports uh

249
00:11:30,000 --> 00:11:33,899
they use something called open ID

250
00:11:31,980 --> 00:11:36,600
connect it's an open Authentication

251
00:11:33,899 --> 00:11:37,640
Protocol that works on top of

252
00:11:36,600 --> 00:11:43,440
um

253
00:11:37,640 --> 00:11:46,380
2.0 framework so uh exam for instance

254
00:11:43,440 --> 00:11:48,600
um if you log into Spotify use in

255
00:11:46,380 --> 00:11:52,200
Facebook you can log into stack also

256
00:11:48,600 --> 00:11:55,500
using Google so that's open uh ID

257
00:11:52,200 --> 00:11:57,120
connect or 2.0 so that's the

258
00:11:55,500 --> 00:12:00,000
Authentication Protocol that's running

259
00:11:57,120 --> 00:12:04,680
behind these solid pods

260
00:12:00,000 --> 00:12:07,320
for authentication purposes and in for

261
00:12:04,680 --> 00:12:10,260
sharing the data it's using something

262
00:12:07,320 --> 00:12:11,700
called web access control ontologies

263
00:12:10,260 --> 00:12:14,600
um

264
00:12:11,700 --> 00:12:18,600
so you can see ah

265
00:12:14,600 --> 00:12:21,240
Imports how do we share the how do we

266
00:12:18,600 --> 00:12:25,019
give access to others to our own data

267
00:12:21,240 --> 00:12:27,420
that's been done using access control

268
00:12:25,019 --> 00:12:31,019
lists so that's this web access control

269
00:12:27,420 --> 00:12:31,620
ontology so you can see in this picture

270
00:12:31,019 --> 00:12:33,839
um

271
00:12:31,620 --> 00:12:38,279
so you have given so they you have given

272
00:12:33,839 --> 00:12:40,680
access to these two people p and that's

273
00:12:38,279 --> 00:12:44,120
line one two three four five

274
00:12:40,680 --> 00:12:47,279
so you have given access to the

275
00:12:44,120 --> 00:12:50,220
filemedical.ttl to those people and what

276
00:12:47,279 --> 00:12:52,500
specific access has been given that's

277
00:12:50,220 --> 00:12:54,959
given in the last line so you have given

278
00:12:52,500 --> 00:12:58,200
control read and write access to those

279
00:12:54,959 --> 00:13:01,860
two people for that medical DOT TTL file

280
00:12:58,200 --> 00:13:03,839
and that's how you usually give so you

281
00:13:01,860 --> 00:13:06,860
can you can give either

282
00:13:03,839 --> 00:13:10,079
one of these three controls uh to to

283
00:13:06,860 --> 00:13:12,779
someone else so

284
00:13:10,079 --> 00:13:16,440
um for transparency you can also have

285
00:13:12,779 --> 00:13:18,420
logs where you can record this access

286
00:13:16,440 --> 00:13:19,079
sharing as well

287
00:13:18,420 --> 00:13:23,300
um

288
00:13:19,079 --> 00:13:23,300
so you can have a look at later stage

289
00:13:25,200 --> 00:13:31,980
and built-in pipelines so when migrating

290
00:13:28,079 --> 00:13:34,440
so this is kind of now we are going into

291
00:13:31,980 --> 00:13:36,779
our project here

292
00:13:34,440 --> 00:13:39,600
um so when migrating from existing

293
00:13:36,779 --> 00:13:42,540
systems uh or coming from accessory

294
00:13:39,600 --> 00:13:45,779
systems to these decentralized databases

295
00:13:42,540 --> 00:13:49,139
these decentralized pods-based systems

296
00:13:45,779 --> 00:13:52,139
one of the challenges we face is that

297
00:13:49,139 --> 00:13:53,399
um we cannot really

298
00:13:52,139 --> 00:13:56,639
um like

299
00:13:53,399 --> 00:13:58,980
fully migrate at once into these Sports

300
00:13:56,639 --> 00:14:02,160
based Concepts right so uh it should be

301
00:13:58,980 --> 00:14:04,620
a step-by-step process a gradual process

302
00:14:02,160 --> 00:14:07,620
and that at some point sports-based

303
00:14:04,620 --> 00:14:10,560
systems need to coexist with the current

304
00:14:07,620 --> 00:14:14,279
trans centralized systems so we need to

305
00:14:10,560 --> 00:14:18,320
build pipelines from and potentially to

306
00:14:14,279 --> 00:14:21,420
these systems from these Sports base

307
00:14:18,320 --> 00:14:24,000
to existing systems using the sport

308
00:14:21,420 --> 00:14:26,720
based systems as well

309
00:14:24,000 --> 00:14:26,720
so

310
00:14:27,060 --> 00:14:32,940
what happens here is we employ use in

311
00:14:31,380 --> 00:14:35,820
our project what we do is we employ

312
00:14:32,940 --> 00:14:38,220
python to build these pipelines so it's

313
00:14:35,820 --> 00:14:41,160
a simple process what we do is we

314
00:14:38,220 --> 00:14:43,980
extract data from existing systems and

315
00:14:41,160 --> 00:14:47,220
then we pre-process data

316
00:14:43,980 --> 00:14:50,360
so in the pre-processing uh what we do

317
00:14:47,220 --> 00:14:53,459
is we sometimes do some data aggregation

318
00:14:50,360 --> 00:14:55,199
maybe do some data encryption data

319
00:14:53,459 --> 00:14:57,720
masking

320
00:14:55,199 --> 00:15:00,180
um things on on top of the extracted

321
00:14:57,720 --> 00:15:04,260
data and then what we do is we transform

322
00:15:00,180 --> 00:15:06,779
data into these Turtle formats so rdf

323
00:15:04,260 --> 00:15:08,760
formats and then we feed that data back

324
00:15:06,779 --> 00:15:11,459
into solid servers

325
00:15:08,760 --> 00:15:14,519
and this can again be done vice versa as

326
00:15:11,459 --> 00:15:16,260
well if needed so we can extract a data

327
00:15:14,519 --> 00:15:19,199
from solid servers do the exact same

328
00:15:16,260 --> 00:15:21,740
thing and feed that data into existing

329
00:15:19,199 --> 00:15:21,740
systems

330
00:15:24,360 --> 00:15:29,639
all right so

331
00:15:26,579 --> 00:15:32,459
um in our project uh so far we have

332
00:15:29,639 --> 00:15:34,260
developed uh so we are mainly working on

333
00:15:32,459 --> 00:15:36,240
something called flutter and Dot for

334
00:15:34,260 --> 00:15:39,060
developing our applications that's in

335
00:15:36,240 --> 00:15:41,699
the front end uh in the backend side we

336
00:15:39,060 --> 00:15:44,519
are using mainly python for instance

337
00:15:41,699 --> 00:15:45,959
building pipelines from these existing

338
00:15:44,519 --> 00:15:48,480
systems

339
00:15:45,959 --> 00:15:50,760
um so flutter is a cross platform app

340
00:15:48,480 --> 00:15:53,820
development technology so the underlying

341
00:15:50,760 --> 00:15:56,820
language has been dot so in this project

342
00:15:53,820 --> 00:15:58,920
we have so far developed three open

343
00:15:56,820 --> 00:16:03,420
source packages

344
00:15:58,920 --> 00:16:05,760
to use with our applications so the

345
00:16:03,420 --> 00:16:10,680
packages are being solid oath solid

346
00:16:05,760 --> 00:16:10,680
encrypt and rdf lib so

347
00:16:10,860 --> 00:16:16,019
solid oath package uh as the name

348
00:16:13,560 --> 00:16:19,980
suggests you can do authentication from

349
00:16:16,019 --> 00:16:22,680
and to pods so you can do uh whenever

350
00:16:19,980 --> 00:16:25,620
you build a flat up uh you know flutter

351
00:16:22,680 --> 00:16:27,839
application you can use this solid

352
00:16:25,620 --> 00:16:30,660
authentication package to do the

353
00:16:27,839 --> 00:16:33,600
authentication into any port out there

354
00:16:30,660 --> 00:16:35,820
right and in the solid encrypt then you

355
00:16:33,600 --> 00:16:39,019
can use this package to encrypt your

356
00:16:35,820 --> 00:16:42,120
data inside your ports at rest

357
00:16:39,019 --> 00:16:43,980
encrypt and decrypt data and the RDA

358
00:16:42,120 --> 00:16:47,339
flip package that's a DOT package for

359
00:16:43,980 --> 00:16:49,220
working with rdf and this package could

360
00:16:47,339 --> 00:16:52,680
simply

361
00:16:49,220 --> 00:16:56,480
extract triplets from the existing

362
00:16:52,680 --> 00:17:00,600
Turtle files in our pods and create this

363
00:16:56,480 --> 00:17:03,720
DF graphs and also

364
00:17:00,600 --> 00:17:07,079
um once we process these graphs we can

365
00:17:03,720 --> 00:17:09,000
export it to a turtle file as well using

366
00:17:07,079 --> 00:17:11,400
this package

367
00:17:09,000 --> 00:17:12,900
so that's a quick introduction to uh

368
00:17:11,400 --> 00:17:16,020
three of the packages that's been

369
00:17:12,900 --> 00:17:17,280
developed on Java project

370
00:17:16,020 --> 00:17:22,280
okay

371
00:17:17,280 --> 00:17:22,280
yes that's where I hand out yes thanks

372
00:17:26,699 --> 00:17:32,059
with the yaraba indigenous community so

373
00:17:29,880 --> 00:17:36,480
their Health Service

374
00:17:32,059 --> 00:17:39,360
Health Service they reached out to us

375
00:17:36,480 --> 00:17:41,520
um they were looking to basically have

376
00:17:39,360 --> 00:17:42,900
better data infrastructure in their I.T

377
00:17:41,520 --> 00:17:45,419
system so they could support

378
00:17:42,900 --> 00:17:47,240
personalized strength-based care

379
00:17:45,419 --> 00:17:50,100
recognizing that in their Community

380
00:17:47,240 --> 00:17:52,140
where they've got a really high disease

381
00:17:50,100 --> 00:17:55,559
burden with diabetes rheumatic heart

382
00:17:52,140 --> 00:17:58,620
disease and a number of other challenges

383
00:17:55,559 --> 00:18:00,840
it's really important to keep people to

384
00:17:58,620 --> 00:18:02,760
bring the care to the patients

385
00:18:00,840 --> 00:18:05,580
and as much as possible keep people

386
00:18:02,760 --> 00:18:08,000
being treated in community so they have

387
00:18:05,580 --> 00:18:10,039
their connection to their country

388
00:18:08,000 --> 00:18:12,299
to their people

389
00:18:10,039 --> 00:18:16,080
which gives them their source of

390
00:18:12,299 --> 00:18:18,720
strength and that uh so they look into

391
00:18:16,080 --> 00:18:20,280
this to have a better system that will

392
00:18:18,720 --> 00:18:21,600
help them to be able to both support

393
00:18:20,280 --> 00:18:23,640
patients

394
00:18:21,600 --> 00:18:26,039
for having better communication with

395
00:18:23,640 --> 00:18:28,500
their doctor around their Integrated

396
00:18:26,039 --> 00:18:31,200
Health and for the doctors to be able to

397
00:18:28,500 --> 00:18:34,880
monitor their patients

398
00:18:31,200 --> 00:18:34,880
um that they're treating in this way

399
00:18:34,919 --> 00:18:38,580
so using solid

400
00:18:36,720 --> 00:18:39,900
eat patient and their doctor has their

401
00:18:38,580 --> 00:18:41,820
own pod

402
00:18:39,900 --> 00:18:43,980
and we're building the health software

403
00:18:41,820 --> 00:18:47,220
where patients can access and control

404
00:18:43,980 --> 00:18:49,080
how the data is shared from their pod

405
00:18:47,220 --> 00:18:50,460
so we've got three pods-based

406
00:18:49,080 --> 00:18:53,160
applications

407
00:18:50,460 --> 00:18:56,760
uh the first is the what we call the

408
00:18:53,160 --> 00:18:58,140
India this is the app for patients which

409
00:18:56,760 --> 00:19:00,600
allows them to access their health

410
00:18:58,140 --> 00:19:03,660
information the advice from their doctor

411
00:19:00,600 --> 00:19:05,820
uh such as the subscriptions

412
00:19:03,660 --> 00:19:08,160
um and and to be able to share their

413
00:19:05,820 --> 00:19:09,660
Integrated Health Data back to their

414
00:19:08,160 --> 00:19:12,120
doctors

415
00:19:09,660 --> 00:19:16,100
control of the sharing then there's a

416
00:19:12,120 --> 00:19:16,100
health worker up because we realized

417
00:19:20,760 --> 00:19:23,940
so we realized that

418
00:19:23,100 --> 00:19:26,220
um

419
00:19:23,940 --> 00:19:28,919
the care coordination teams that going

420
00:19:26,220 --> 00:19:31,559
out to patients in the community

421
00:19:28,919 --> 00:19:33,059
they're actually using paper and Pen to

422
00:19:31,559 --> 00:19:34,620
do much of their recording and then

423
00:19:33,059 --> 00:19:36,000
having the load of bringing that back to

424
00:19:34,620 --> 00:19:37,620
the clinic and having to put it back

425
00:19:36,000 --> 00:19:40,140
into their their patient information

426
00:19:37,620 --> 00:19:42,480
management system so this allows them to

427
00:19:40,140 --> 00:19:44,820
have the patient data with them on the

428
00:19:42,480 --> 00:19:46,799
Fly throughout the community be able to

429
00:19:44,820 --> 00:19:47,700
update that and have it sync back to

430
00:19:46,799 --> 00:19:49,919
their system

431
00:19:47,700 --> 00:19:52,520
and then the third app is the granny up

432
00:19:49,919 --> 00:19:56,880
uh and this is the one that's gonna

433
00:19:52,520 --> 00:19:59,039
provide the analytics uh to show how

434
00:19:56,880 --> 00:20:02,340
patients who are

435
00:19:59,039 --> 00:20:04,880
being treated in this way uh performing

436
00:20:02,340 --> 00:20:04,880
in the community

437
00:20:05,160 --> 00:20:09,780
so we're going to set up yaraba solid

438
00:20:06,900 --> 00:20:11,900
server where each patient has their pod

439
00:20:09,780 --> 00:20:14,640
and they're each individually encrypted

440
00:20:11,900 --> 00:20:16,919
and those pods can be hosted on any

441
00:20:14,640 --> 00:20:18,960
solid server

442
00:20:16,919 --> 00:20:21,360
in the India the patients have their

443
00:20:18,960 --> 00:20:23,760
granular control over what data they

444
00:20:21,360 --> 00:20:25,679
share or decide not to share

445
00:20:23,760 --> 00:20:28,020
and that also means that they can

446
00:20:25,679 --> 00:20:30,539
delegate control perhaps to a trusted

447
00:20:28,020 --> 00:20:32,460
other person

448
00:20:30,539 --> 00:20:35,340
and those Access Control lists that you

449
00:20:32,460 --> 00:20:38,280
saw there they control access to each of

450
00:20:35,340 --> 00:20:40,320
the rdf resources in their pod

451
00:20:38,280 --> 00:20:43,260
they can view the access that they have

452
00:20:40,320 --> 00:20:46,520
given and change access to stop sharing

453
00:20:43,260 --> 00:20:46,520
to a particular recipient

454
00:20:47,520 --> 00:20:52,880
store additional data and share that to

455
00:20:50,280 --> 00:20:52,880
another person

456
00:20:53,820 --> 00:20:57,720
so we're beginning a trial of this

457
00:20:55,679 --> 00:21:00,539
pods-based up system with the patients

458
00:20:57,720 --> 00:21:02,940
working with a group of low-risk

459
00:21:00,539 --> 00:21:04,620
diabetes patients and being recruited to

460
00:21:02,940 --> 00:21:07,039
the trial and these will have the

461
00:21:04,620 --> 00:21:09,799
opportunity to engage

462
00:21:07,039 --> 00:21:12,840
with the medical staff and with us

463
00:21:09,799 --> 00:21:15,120
around the design and development of

464
00:21:12,840 --> 00:21:16,740
these applications for them so there'll

465
00:21:15,120 --> 00:21:18,179
be a number of yawning sessions where

466
00:21:16,740 --> 00:21:19,260
we're developing the data governance

467
00:21:18,179 --> 00:21:22,740
framework

468
00:21:19,260 --> 00:21:25,080
uh the the different leaders across the

469
00:21:22,740 --> 00:21:27,780
community want to have for the patients

470
00:21:25,080 --> 00:21:29,340
working with representatives and then

471
00:21:27,780 --> 00:21:31,200
that'll be implemented in the consent

472
00:21:29,340 --> 00:21:33,419
model within the pods

473
00:21:31,200 --> 00:21:36,299
and then working with those patients who

474
00:21:33,419 --> 00:21:39,000
are able to come to those sessions to to

475
00:21:36,299 --> 00:21:40,700
better improve how we can give them

476
00:21:39,000 --> 00:21:44,340
access

477
00:21:40,700 --> 00:21:46,860
and make it transparent uh the consent

478
00:21:44,340 --> 00:21:49,440
that they are able to control

479
00:21:46,860 --> 00:21:50,760
and at the end of the trial we envisaged

480
00:21:49,440 --> 00:21:52,020
that there would actually be multiple

481
00:21:50,760 --> 00:21:54,059
Community groups that would have

482
00:21:52,020 --> 00:21:56,940
pods-based applications

483
00:21:54,059 --> 00:21:58,500
and then this would allow

484
00:21:56,940 --> 00:22:01,140
patients who are attending different

485
00:21:58,500 --> 00:22:04,799
activities you know whether it's men

486
00:22:01,140 --> 00:22:07,559
shared the Arts Community or the footy

487
00:22:04,799 --> 00:22:10,679
Club to then have their attendance in

488
00:22:07,559 --> 00:22:12,600
those activities immediately ingested

489
00:22:10,679 --> 00:22:14,400
within their pod

490
00:22:12,600 --> 00:22:15,659
and this is the kind of information that

491
00:22:14,400 --> 00:22:18,059
the doctors

492
00:22:15,659 --> 00:22:20,100
believe is really important to the

493
00:22:18,059 --> 00:22:21,960
cultural and Integrated Health of their

494
00:22:20,100 --> 00:22:24,419
patients

495
00:22:21,960 --> 00:22:26,539
so using pods we're able to provide that

496
00:22:24,419 --> 00:22:29,400
infrastructure for

497
00:22:26,539 --> 00:22:30,620
distributed or individually governed

498
00:22:29,400 --> 00:22:33,419
data

499
00:22:30,620 --> 00:22:35,220
which can be just within communities or

500
00:22:33,419 --> 00:22:37,980
it can be extended outside communities

501
00:22:35,220 --> 00:22:40,679
to other entities that have pods to be

502
00:22:37,980 --> 00:22:42,419
able to receive data using solid

503
00:22:40,679 --> 00:22:44,940
technology

504
00:22:42,419 --> 00:22:47,100
and each person can then be in control

505
00:22:44,940 --> 00:22:51,240
of the data that they share

506
00:22:47,100 --> 00:22:55,039
and confident that it is encrypted and

507
00:22:51,240 --> 00:22:55,039
safe within those solid servers

508
00:22:56,400 --> 00:23:02,520
so this raises a number of challenges

509
00:23:00,539 --> 00:23:05,159
we briefly mentioned you know the

510
00:23:02,520 --> 00:23:08,039
challenge of setting up the pipelines

511
00:23:05,159 --> 00:23:10,260
and migration to those existing systems

512
00:23:08,039 --> 00:23:12,299
because moving into a pods-based

513
00:23:10,260 --> 00:23:13,620
architecture of data the data is

514
00:23:12,299 --> 00:23:15,480
effectively stored like a Knowledge

515
00:23:13,620 --> 00:23:17,460
Graph and so there's that transposition

516
00:23:15,480 --> 00:23:19,860
from a relational database

517
00:23:17,460 --> 00:23:23,220
or other type of database in into the

518
00:23:19,860 --> 00:23:25,220
graph rdf format

519
00:23:23,220 --> 00:23:28,440
but once it is in that format

520
00:23:25,220 --> 00:23:30,679
standardized rdf formats then any other

521
00:23:28,440 --> 00:23:32,280
external pods-based application

522
00:23:30,679 --> 00:23:35,340
could

523
00:23:32,280 --> 00:23:37,440
through the access control list request

524
00:23:35,340 --> 00:23:39,179
access and basically offer that software

525
00:23:37,440 --> 00:23:42,020
to be able to connect with a person's

526
00:23:39,179 --> 00:23:42,020
data in their pod

527
00:23:42,720 --> 00:23:47,100
then there's also the aspect of data

528
00:23:44,580 --> 00:23:48,900
security so while the data is relatively

529
00:23:47,100 --> 00:23:51,240
safe because it is only decrypted on

530
00:23:48,900 --> 00:23:52,860
device that still does open the

531
00:23:51,240 --> 00:23:55,020
potential for back doors on those

532
00:23:52,860 --> 00:23:57,740
devices

533
00:23:55,020 --> 00:23:57,740
and

534
00:23:58,620 --> 00:24:03,780
we're in this future environment where

535
00:24:01,500 --> 00:24:05,340
privacy has become much more important

536
00:24:03,780 --> 00:24:07,440
people becoming much more aware of

537
00:24:05,340 --> 00:24:09,740
privacy we have legislation changing

538
00:24:07,440 --> 00:24:12,960
throughout the world around privacy

539
00:24:09,740 --> 00:24:15,120
you know being led very much by the gdpr

540
00:24:12,960 --> 00:24:16,460
in Europe

541
00:24:15,120 --> 00:24:19,799
however

542
00:24:16,460 --> 00:24:21,240
implementing those rights is really

543
00:24:19,799 --> 00:24:22,679
really challenging

544
00:24:21,240 --> 00:24:23,760
so one of those is the right to be

545
00:24:22,679 --> 00:24:26,100
forgotten

546
00:24:23,760 --> 00:24:27,480
with pods-based data storage you can

547
00:24:26,100 --> 00:24:29,280
turn off access

548
00:24:27,480 --> 00:24:31,380
so perhaps I've been giving my birth

549
00:24:29,280 --> 00:24:34,140
date as a piece of information to

550
00:24:31,380 --> 00:24:36,000
government agencies and then I say no I

551
00:24:34,140 --> 00:24:40,020
no longer want to do that

552
00:24:36,000 --> 00:24:40,020
I've lost faith in them and

553
00:24:41,000 --> 00:24:45,360
but then you don't know whether or not

554
00:24:43,500 --> 00:24:47,820
it is deleted

555
00:24:45,360 --> 00:24:51,120
so there is at the moment it's still

556
00:24:47,820 --> 00:24:53,100
dependent on this aspect of trust even

557
00:24:51,120 --> 00:24:55,760
though we do want to move to a trust no

558
00:24:53,100 --> 00:24:55,760
one environment

559
00:24:57,360 --> 00:25:01,620
so and finally

560
00:24:59,640 --> 00:25:04,020
um consent models for data sharing need

561
00:25:01,620 --> 00:25:06,059
to be simple to understand and use as

562
00:25:04,020 --> 00:25:08,340
well as being feasible to build consent

563
00:25:06,059 --> 00:25:10,140
is not consent if the individual doesn't

564
00:25:08,340 --> 00:25:11,700
understand what they have consented or

565
00:25:10,140 --> 00:25:14,480
they're not able to review or change

566
00:25:11,700 --> 00:25:14,480
that consent

567
00:25:16,740 --> 00:25:21,419
so working with the indigenous

568
00:25:19,620 --> 00:25:24,360
communities that have a very strong

569
00:25:21,419 --> 00:25:25,440
understanding of

570
00:25:24,360 --> 00:25:27,659
um

571
00:25:25,440 --> 00:25:30,360
of data governance

572
00:25:27,659 --> 00:25:32,940
we're able to benefit from that strength

573
00:25:30,360 --> 00:25:35,640
of history that they have there

574
00:25:32,940 --> 00:25:38,700
and

575
00:25:35,640 --> 00:25:40,500
their understanding around you know for

576
00:25:38,700 --> 00:25:43,260
instance that access to data is to be

577
00:25:40,500 --> 00:25:44,700
earned and we can design and build the

578
00:25:43,260 --> 00:25:46,740
consent model that supports the

579
00:25:44,700 --> 00:25:49,080
communities to have the kind of access

580
00:25:46,740 --> 00:25:51,559
and control over their data that they

581
00:25:49,080 --> 00:25:51,559
are seeking

582
00:25:52,080 --> 00:25:56,580
so with that I just want to say thank

583
00:25:53,640 --> 00:25:58,799
you thank you to you all for coming to

584
00:25:56,580 --> 00:26:01,919
hear about building with pods and also

585
00:25:58,799 --> 00:26:03,539
to our team so this work

586
00:26:01,919 --> 00:26:04,980
um you know wouldn't wouldn't be

587
00:26:03,539 --> 00:26:09,419
possible without the trust of the

588
00:26:04,980 --> 00:26:11,580
governor Health Service up in yaraba

589
00:26:09,419 --> 00:26:13,799
um and without our wonderful team led by

590
00:26:11,580 --> 00:26:15,960
Professor Graham Williams in the school

591
00:26:13,799 --> 00:26:16,620
of computing at Anu

592
00:26:15,960 --> 00:26:19,500
um

593
00:26:16,620 --> 00:26:22,980
some of our researchers Sergio Rodrigo

594
00:26:19,500 --> 00:26:26,400
Mendez and our software Engineers Nina

595
00:26:22,980 --> 00:26:28,940
Bart and saman Shu and Nunez with us

596
00:26:26,400 --> 00:26:28,940
here today

597
00:26:31,799 --> 00:26:35,950
thank you

598
00:26:32,810 --> 00:26:35,950
[Music]

599
00:26:37,500 --> 00:26:42,720
masks and earrings

600
00:26:39,600 --> 00:26:45,059
and mics and earrings apparently

601
00:26:42,720 --> 00:26:46,320
uh does anyone have any questions and

602
00:26:45,059 --> 00:26:49,580
you're gonna have to make yourself clear

603
00:26:46,320 --> 00:26:49,580
because I can't see much

604
00:26:52,580 --> 00:26:57,659
uh is there any expiry based ACL

605
00:26:55,679 --> 00:27:01,080
structure so you might want something

606
00:26:57,659 --> 00:27:02,640
where you don't want somebody to like

607
00:27:01,080 --> 00:27:04,559
instead of having to revoke it is there

608
00:27:02,640 --> 00:27:06,240
a way to have it that it you can set the

609
00:27:04,559 --> 00:27:09,299
expiry on it

610
00:27:06,240 --> 00:27:12,059
all right now we only have revoking

611
00:27:09,299 --> 00:27:14,460
access in the solid pods technology but

612
00:27:12,059 --> 00:27:18,799
we can build on top of that to have

613
00:27:14,460 --> 00:27:18,799
expiry this you know his students

614
00:27:31,500 --> 00:27:36,000
um migrating to like a semantic

615
00:27:33,539 --> 00:27:36,960
web-based technology is the learning

616
00:27:36,000 --> 00:27:38,340
challenge

617
00:27:36,960 --> 00:27:40,320
um are there communities that you

618
00:27:38,340 --> 00:27:42,600
participate with to like kind of start

619
00:27:40,320 --> 00:27:44,039
building up that knowledge skills that

620
00:27:42,600 --> 00:27:46,200
are going to be required to get to

621
00:27:44,039 --> 00:27:48,620
semantic web-based Technologies in

622
00:27:46,200 --> 00:27:48,620
Australia

623
00:27:49,260 --> 00:27:54,480
yeah okay uh so we got quite a lot of

624
00:27:52,020 --> 00:27:57,480
specialty around semantic Technologies

625
00:27:54,480 --> 00:28:00,120
in the school of community computing

626
00:27:57,480 --> 00:28:02,520
and through their involvement and their

627
00:28:00,120 --> 00:28:05,279
connections to those communities is is

628
00:28:02,520 --> 00:28:09,120
really how we're connected there but

629
00:28:05,279 --> 00:28:10,799
certainly we are seeing this movement to

630
00:28:09,120 --> 00:28:12,080
setting up communities working with

631
00:28:10,799 --> 00:28:15,120
solid

632
00:28:12,080 --> 00:28:18,360
which a large part of that is recognized

633
00:28:15,120 --> 00:28:21,779
that needing to drive the um the

634
00:28:18,360 --> 00:28:25,200
development of standards in how how we

635
00:28:21,779 --> 00:28:28,039
structure and store data in in rdf and

636
00:28:25,200 --> 00:28:28,039
triple format

637
00:28:28,980 --> 00:28:32,460
we're sorry to cut off the other people

638
00:28:30,480 --> 00:28:34,380
but you probably noticed we're a little

639
00:28:32,460 --> 00:28:37,799
bit behind

640
00:28:34,380 --> 00:28:41,460
so we'll defer hallway questions

641
00:28:37,799 --> 00:28:43,980
go find uh Anushka or Jess out in the

642
00:28:41,460 --> 00:28:46,799
hall and we wanted to say thank you very

643
00:28:43,980 --> 00:28:48,539
much to both of you and redwin has a

644
00:28:46,799 --> 00:28:51,059
little gift for you as well before you

645
00:28:48,539 --> 00:28:55,560
bail on the room uh we'll see you all

646
00:28:51,059 --> 00:28:59,099
back here at 11 for our next talk

647
00:28:55,560 --> 00:28:59,099
[Applause]