0:00:07.278,0:00:11.778 [inaudible] and I [br]have an effort called WikiLoop, 0:00:11.778,0:00:15.368 and this is what I'm going [br]to introduce to you about. 0:00:15.728,0:00:22.604 We have presented WikiLoop, the idea, [br]to several Wikimedia related conferences. 0:00:22.604,0:00:25.017 How many of you have heard [br]about WikiLoop before? 0:00:26.020,0:00:27.040 Thanks. 0:00:27.040,0:00:31.014 And how many of you have interacted [br]with the datasets and toolings 0:00:31.014,0:00:32.664 that we provided before? 0:00:33.308,0:00:36.870 Okay, fairly new. [br]So this will be mostly an introduction. 0:00:36.870,0:00:42.008 So we would like to tell you [br]why we start this initiative 0:00:42.008,0:00:44.148 and what it intends to do, 0:00:44.148,0:00:48.803 and how you can get involved [br]or what it will go for. 0:00:50.390,0:00:53.810 So, to begin with, [br]we would like to give you an example. 0:00:53.810,0:00:58.409 This is a vandalism [br]that happened in Italian... 0:01:00.621,0:01:03.623 that happened in Italy Wikipedia. 0:01:04.142,0:01:06.935 I know that most people here [br]are interested in Wikidata. 0:01:06.935,0:01:09.780 I will tell you why this is relevant too. 0:01:10.137,0:01:11.879 So basically what we found is 0:01:11.879,0:01:15.970 that someone vandalized [br]Wikipedia on Italian 0:01:15.970,0:01:20.590 and says, "Bezos who cannot afford a car." 0:01:20.809,0:01:22.666 And this is an interesting question, 0:01:23.799,0:01:28.379 if you think about it,[br]this is blatant obvious vandalism 0:01:28.379,0:01:33.412 but when it comes to machines[br]and algorithms 0:01:33.412,0:01:37.881 which find to detect vandalism [br]and avoid serving users the information, 0:01:38.309,0:01:41.989 how can computer understand [br]this kind of information, 0:01:41.989,0:01:43.286 like it would be... 0:01:46.869,0:01:49.180 we realize that sometimes [br]there are limitations 0:01:49.180,0:01:54.083 of how far algorithms can go [br]and machine can go. 0:01:54.931,0:01:57.666 Another example here is let's say, 0:01:57.666,0:02:02.044 there is a word or label, [br]or a category on Wikipedia says, 0:02:02.044,0:02:06.077 someone, a person,[br]is a Christian scientist. 0:02:06.077,0:02:09.627 Now, given this label, [br]what facts do you come up with 0:02:09.627,0:02:13.815 like what would you infer [br]from this category? 0:02:14.205,0:02:18.586 Do you think it would be a "Christian"[br]or do you think it would be a "scientist"? 0:02:18.981,0:02:21.621 In this specific case--[br]it does not apply everywhere-- 0:02:21.621,0:02:23.481 but it this specific case, 0:02:23.481,0:02:26.991 there is a religion [br]called "Christian Science," 0:02:26.991,0:02:30.199 and people who hold that belief[br]is called "Christian Scientist." 0:02:31.549,0:02:34.891 And, again, for machines, [br]how can we know, like 0:02:36.272,0:02:40.392 even if many people here are big [fan] 0:02:40.392,0:02:45.242 that's the better we make our data [br]a knowledge machine-friendly 0:02:45.459,0:02:51.709 the easier we can work and improve [br]the overall knowledge accessibility 0:02:51.709,0:02:54.139 and contribute together 0:02:54.139,0:02:55.589 but there is always things 0:02:55.589,0:02:58.449 that we believe [br]that machine has restrictions. 0:03:00.136,0:03:04.479 So all in all, we start to realize 0:03:04.479,0:03:08.307 that coming from Internet companies 0:03:08.307,0:03:10.690 who have a strong belief [br]of our technology 0:03:10.690,0:03:12.571 and what machine can do, 0:03:12.571,0:03:16.222 there is always a gap [br]or there is always something 0:03:16.222,0:03:18.992 that we would need to rely on human being 0:03:18.992,0:03:22.442 and more, we would need[br]to rely on communities 0:03:22.753,0:03:28.383 who are actively contributing,[br]who are doing the peer reviews to our... 0:03:28.383,0:03:30.163 collaborating with each other. 0:03:30.163,0:03:36.082 So this is a picture [br]about the background effort of WikiLoop. 0:03:36.595,0:03:39.945 For the human being, [br]they have the knowledge, 0:03:40.485,0:03:46.205 we have our domain expertize [br]and we can crosscheck each other 0:03:46.205,0:03:48.503 but we just have that enough time. 0:03:49.333,0:03:52.803 And there are many things [br]that machine can empower this 0:03:52.803,0:03:56.123 but there is restrictions there. 0:03:56.123,0:03:58.643 So the goal is to empower 0:03:58.643,0:04:03.039 or improve the productivity [br]of human editors. 0:04:03.039,0:04:08.633 But also the other side of the formula [br]is we want to loop that back 0:04:08.634,0:04:13.234 to the research and the academic efforts 0:04:13.234,0:04:17.312 that improve how machine [br]can help in these cases. 0:04:17.875,0:04:22.580 So by raise of hand, [br]how many of you have used Google? 0:04:23.870,0:04:25.090 Thank you. 0:04:25.090,0:04:26.380 And how many of you 0:04:26.900,0:04:31.455 think that companies like Google [br]and other big knowledge companies 0:04:31.455,0:04:34.202 should contribute more [br]to the knowledge world? 0:04:35.881,0:04:37.707 So what happens is that... 0:04:37.707,0:04:42.157 we all know that our mission at Google [br]or other similar companies-- 0:04:42.157,0:04:47.647 we have a strong background[br]of leveraging the open knowledge world, 0:04:48.347,0:04:50.107 like for Google specific case 0:04:50.107,0:04:52.740 it's like organize [br]the world's information. 0:04:52.740,0:04:55.059 So we help disseminate the information, 0:04:56.207,0:04:59.996 which in one sense that helps [br]the mission of this movement. 0:04:59.996,0:05:06.358 But only every once a while [br]we have sporadic help 0:05:07.864,0:05:12.103 trying to donate knowledge [br]and datasets, and tools, 0:05:12.103,0:05:16.223 and we want to see [br]if we can make this sustainable, 0:05:18.323,0:05:21.424 both in the technical sense 0:05:21.424,0:05:23.234 and also in the business sense. 0:05:24.943,0:05:29.639 So this is like [br]a one-sentence introduction. 0:05:29.639,0:05:34.885 We want WikiLoop [br]to become an umbrella program 0:05:34.885,0:05:37.084 for a series of technical projects 0:05:37.084,0:05:39.632 intended to contribute [br]datasets and toolings 0:05:39.632,0:05:44.734 and hopefully make this a community effort[br]with participation of 0:05:44.734,0:05:50.154 other likeminded people, [br]partners and institutions 0:05:50.154,0:05:52.410 to join with this effort. 0:05:52.410,0:05:56.204 There are several projects [br]that we think would be a good fit, 0:05:56.204,0:05:59.204 and these are the criteria. 0:05:59.204,0:06:04.281 First of all, the idea is [br]that it needs to be source improvements 0:06:04.281,0:06:07.251 or source improvements[br]by and large is a good fit, 0:06:07.251,0:06:10.801 and also the second thing [br]that companies like us 0:06:10.801,0:06:13.941 really cannot do very well by ourself 0:06:13.941,0:06:17.691 is to maximize the neutrality,[br]to avoid picking sides 0:06:17.691,0:06:21.611 on the controversies, [br]decisions or discussions 0:06:21.611,0:06:26.945 and another thing is that to make this [br]in the long-term sustainability 0:06:26.945,0:06:31.705 and to keep it[br]being supported by this industry. 0:06:31.705,0:06:35.017 We want to see [br]the productivity, the scalability 0:06:35.017,0:06:37.632 of our contribution and efforts. 0:06:38.444,0:06:41.078 To explain a little bit more... 0:06:41.584,0:06:43.570 We always look trying to extract... 0:06:43.570,0:06:47.061 for example, we are trying [br]to extract facts from Wikipedia. 0:06:47.417,0:06:52.539 And while we can do [br]several separations, 0:06:52.539,0:06:55.704 we're labeling, fairly well, 0:06:56.315,0:06:59.915 up to certain point [br]the bottleneck is no longer 0:06:59.915,0:07:02.475 how good the machine, [br]the algorithm can reach 0:07:02.475,0:07:06.117 but sometimes [br]there is a noise in the source, 0:07:06.117,0:07:10.917 and if we do not remove the source 0:07:10.917,0:07:13.624 or minimize the source noise there, 0:07:13.624,0:07:15.634 that's how far the machine can go. 0:07:15.634,0:07:18.024 So that's the first criteria. 0:07:18.024,0:07:19.383 And the second criteria is, 0:07:19.383,0:07:24.492 we don't want to get to be seen as buyers[br]or introduce potential buyers. 0:07:24.492,0:07:29.822 We want to rely on [br]governance that is peer reviewed 0:07:29.822,0:07:32.686 and that is done by the community 0:07:32.686,0:07:36.570 so that we can avoid picking sides [br]in the controversy questions. 0:07:37.319,0:07:40.809 And the third thing [br]which probably not so intuitive 0:07:40.809,0:07:43.309 but this is the kind of... [br]I would like... 0:07:43.309,0:07:48.039 Let me give you an example [br]of the projects we have in mind. 0:07:48.435,0:07:51.665 Let's say there are smaller, [br]minority language there. 0:07:51.665,0:07:55.940 I have heard a very good talk [br]earlier this morning. 0:07:55.940,0:07:58.460 But one idea we have here is, 0:07:58.460,0:08:02.050 let's say you are a minority [br]language contributor, very active, 0:08:02.050,0:08:07.063 and you want to advocate for your culture [br]and supporting your knowledge creation. 0:08:07.607,0:08:11.747 But because companies like Google[br]or other consumer company, 0:08:11.747,0:08:14.795 they have a bar [br]for releasing a translation, 0:08:14.795,0:08:16.165 to make it available. 0:08:16.165,0:08:18.837 They want the precision to be high enough 0:08:18.837,0:08:21.594 so that they can use it to serve users. 0:08:22.568,0:08:26.568 But maybe internally they have AI modules [br]that are experimenting, 0:08:26.568,0:08:28.914 not good enough to the bar 0:08:28.914,0:08:31.494 because lack of training data, 0:08:32.734,0:08:34.834 so the translation is not available. 0:08:34.834,0:08:38.080 But the community is doing [br]the translation by hand anyway. 0:08:39.160,0:08:41.170 Now, one of the things we are thinking of, 0:08:41.170,0:08:45.170 if we can provide [br]some of this experimental thing 0:08:45.170,0:08:47.660 that is not good enough [br]to serve general user purpose 0:08:47.660,0:08:50.350 but still good for the community 0:08:50.350,0:08:53.558 and somewhat improving the productivity, 0:08:53.811,0:08:55.731 it would be able to 0:08:55.731,0:09:01.381 one, improve the speed of how well [br]a community can contribute, 0:09:01.381,0:09:06.231 and second, what a community is creating [br]anyway can come back as a training data 0:09:06.231,0:09:08.881 that keeps bootstrapping the machines. 0:09:10.376,0:09:15.406 So over time by this effort [br]we hope to generate a model 0:09:15.673,0:09:19.463 that both helps [br]the human being, the editors, 0:09:19.463,0:09:22.246 but also helps the research 0:09:22.246,0:09:26.765 that improves the AI and other approaches. 0:09:28.489,0:09:31.549 And this is a big overview [br]of a few projects 0:09:31.549,0:09:33.509 we are going to introduce. 0:09:33.509,0:09:36.539 Due to the time limitation[br]I will feature a few. 0:09:36.539,0:09:41.492 The WikiLoop Game, which you can look up, 0:09:41.492,0:09:46.732 is one that we leveraged a platform 0:09:46.732,0:09:50.057 created by Magnus called Wikidata Game. 0:09:50.057,0:09:54.847 We provide several datasets there [br]to be played, to be introduced 0:09:54.847,0:09:56.677 and commit to the Wikidata 0:09:56.677,0:09:58.867 but by the human review. 0:09:59.727,0:10:03.947 And Google doesn't get [br]to contribute data directly 0:10:03.947,0:10:06.257 to Wikipedia or Wikidata 0:10:06.257,0:10:12.269 but having someone who is reviewing it [br]as non-biased individuals to do so. 0:10:12.550,0:10:16.620 And the second one I'm going to feature [br]is WikiLoop Battlefield, 0:10:16.620,0:10:21.420 the one that you have seen just now [br]as a counter-vandalism platform, 0:10:21.420,0:10:25.629 and this one also features [br]the same criteria 0:10:25.629,0:10:28.029 of source improvements, 0:10:29.918,0:10:33.328 of how it can empower machines 0:10:33.328,0:10:38.794 by looping back to the training data 0:10:38.794,0:10:43.064 and also how it avoids companies like us 0:10:43.064,0:10:48.526 to pick sides allowing way to rely [br]on the community's assessment. 0:10:48.526,0:10:53.517 And the third one is CitePool,[br]which is creating... 0:10:53.517,0:10:58.469 we're trying to help creating [br]citation candidate pool 0:10:58.469,0:11:02.731 to improve the productivity of people [br]who want to add citation 0:11:02.731,0:11:04.721 but also see if we can make that 0:11:04.721,0:11:09.569 into a training data [br]accessible to researchers. 0:11:10.010,0:11:13.120 So let me use WikiLoop Battlefield [br]as an example. 0:11:13.120,0:11:18.427 If you have... try it on your phone-- [br]battlefield.wikiloop.org. 0:11:18.427,0:11:21.575 By the way, I want to highlight, [br]the name is subject to change 0:11:21.575,0:11:25.870 because some friendly community members [br]have come to me and suggest 0:11:25.870,0:11:32.224 that Battlefield might not be [br]the best name for a project 0:11:32.224,0:11:34.653 serving the Wikimedia movement. 0:11:34.952,0:11:39.542 So if you don't like this name,[br]come join us in the discussion, 0:11:39.542,0:11:40.984 provide your suggestion, 0:11:40.984,0:11:44.499 we will be very happy [br]to converge to a name 0:11:44.499,0:11:48.111 that has community consensus [br]and popularity. 0:11:48.244,0:11:51.166 But let's use that as a placeholder here. 0:11:52.885,0:11:56.500 I don't need to introduce [br]to this group of people 0:11:56.500,0:11:59.097 about the typical vandalism workflow 0:11:59.820,0:12:03.400 but if you have already... 0:12:04.934,0:12:08.886 trying to conduct [br]some counter-vandalism activity, 0:12:08.886,0:12:11.566 you might know that it's not very trivial. 0:12:11.566,0:12:16.413 How many of you have seen vandalism [br]on Wikipedia and Wikidata? 0:12:16.992,0:12:22.329 Okay, how many of you [br]have reverted, by hand, some of them? 0:12:22.890,0:12:27.680 How many of you have used certain tools [br]or go ahead and find certain tools 0:12:27.680,0:12:30.875 to patrol or revert vandalism? 0:12:31.407,0:12:32.497 Okay. 0:12:33.474,0:12:36.124 Cool, this is [br]the highest density of people 0:12:36.124,0:12:41.264 who have tried to revert vandalism 0:12:41.264,0:12:43.625 that I have spoken to before. 0:12:44.336,0:12:48.756 So maybe some of you have been [br]very comfortably doing that 0:12:48.756,0:12:52.966 but for me as someone [br]who started editing actively 0:12:53.808,0:12:57.348 only since like three years ago 0:12:57.562,0:13:03.439 and who only started to be very serous [br]doing vandalism detection and patrolling 0:13:03.879,0:13:06.191 only since about last year 0:13:06.428,0:13:10.836 I found that doing so is not super easy 0:13:10.836,0:13:14.161 on the world of Wikimedia movement. 0:13:15.080,0:13:21.761 If we look at the existing alternatives 0:13:21.761,0:13:25.761 there are tools that is built [br]featuring desktops, 0:13:25.761,0:13:30.748 there are tools that is relying [br]on users who have rollback permissions, 0:13:30.748,0:13:33.976 which itself is a big barrier to get. 0:13:35.248,0:13:39.097 We want to make this [br]a super easy to use platform 0:13:39.097,0:13:41.637 for all the three roles. 0:13:41.637,0:13:46.017 The first one is user, reviewer or editor,[br]whatever you call it. 0:13:46.612,0:13:48.460 The second one is researcher 0:13:48.460,0:13:52.982 who is trying to create [br]vandalism detecting algorithms or systems. 0:13:52.982,0:13:54.732 And the third one is developers 0:13:54.732,0:13:59.573 who is trying to improve [br]this WikiLoop Battlefield tooling. 0:13:59.573,0:14:02.241 We want it to be [br]super easy for user to use. 0:14:02.241,0:14:04.970 You can you pull up your phone, [br]you don't have to install it, 0:14:04.970,0:14:07.168 you can do in on your laptop. 0:14:07.168,0:14:10.170 And we also want [br]to lower a barrier to review. 0:14:10.170,0:14:16.650 The reason why other tools [br]are trying to limit the access to the tool 0:14:16.650,0:14:22.250 is because there needs to be [br]a base trust level for people to use them. 0:14:22.250,0:14:26.634 You don't want someone [br]to come to a counter-vandalism tool 0:14:26.634,0:14:28.226 to vandalize itself. 0:14:29.259,0:14:32.479 So what we are trying to do is that, 0:14:32.479,0:14:34.489 to begin with, we want [br]to make it super easy 0:14:34.489,0:14:39.522 but also we want to allow multiple people[br]to label the same thing. 0:14:39.968,0:14:42.258 Also we want to make it super convenient 0:14:42.258,0:14:48.240 to see the [inaudible], [br]to see other label, and all in real time. 0:14:48.438,0:14:52.317 We also want to make it [br]for researchers super easy to use. 0:14:52.317,0:14:55.227 By one click you can download the labeling 0:14:55.227,0:15:01.356 and maybe start play with the data [br]and see how it fits in your model. 0:15:01.502,0:15:06.129 And we provide APIs [br]that have access to real time data. 0:15:06.758,0:15:10.448 And for the developer [br]we make it very easy to pick up-- 0:15:10.448,0:15:15.433 we have one click-- [br]you can deploy your trial instances, 0:15:15.433,0:15:16.726 things like that. 0:15:17.100,0:15:20.820 This is an example [br]about building projects 0:15:20.820,0:15:23.191 for umbrella like WikiLoop. 0:15:23.191,0:15:27.637 We want to make sure [br]the community trust comes the first. 0:15:27.947,0:15:31.336 We usually need to make it [br]open source the best. 0:15:31.800,0:15:37.478 And we want to avoid proprietary tech,[br]we want to avoid tech lock-down, 0:15:37.778,0:15:42.999 and we rely on community approval [br]for certain features. 0:15:44.366,0:15:49.474 And if you have seen this--[br]this is the components that we rely on-- 0:15:49.474,0:15:56.207 still very early stage but you get [br]the principles behind the design. 0:15:56.438,0:16:00.288 So what's next, we are trying [br]to grow our usage. 0:16:00.288,0:16:02.458 Hopefully you can try it out by yourself 0:16:02.458,0:16:06.726 and promise to me [br]that you don't click on the login. 0:16:07.782,0:16:09.132 There is a login button-- 0:16:09.132,0:16:10.452 there will be some good features 0:16:10.452,0:16:13.292 that make it super easy [br]to even revert something. 0:16:13.292,0:16:15.452 Currently it's still a jump to revert. 0:16:16.714,0:16:18.444 But we are building features, 0:16:18.444,0:16:23.954 and we are also trying [br]to let you choose some categories 0:16:23.954,0:16:26.656 or the watchlist [br]that you will be watching 0:16:26.656,0:16:31.366 and the one that you care about to patrol. 0:16:31.775,0:16:38.069 And also if you are researchers [br]while doing related vandalism detection, 0:16:38.362,0:16:41.580 try our data and give us feedback. 0:16:44.411,0:16:47.181 And I will go through quickly [br]about a few other projects 0:16:47.181,0:16:48.731 that we are featuring here 0:16:48.731,0:16:52.171 and we will look for questions [br]and feedback from you 0:16:52.171,0:16:57.976 about what we think [br]and what you think should be there 0:16:57.976,0:17:01.550 or how we should fix things [br]if it doesn't work right. 0:17:01.843,0:17:06.163 Wikidata Game is a platform [br]built by a community member Magnus, 0:17:06.163,0:17:08.913 a celebrity in this community, I think. 0:17:09.891,0:17:13.371 And by showing this [br]we are providing datasets 0:17:13.371,0:17:19.748 but we also want to let people know[br]that we are not reinventing the wheels, 0:17:19.748,0:17:21.368 that we are not trying to... 0:17:21.368,0:17:24.168 When we come up with some idea, [br]we look into with community 0:17:24.168,0:17:27.028 and see if there is [br]existing tools that's there 0:17:27.028,0:17:30.198 and how we can be [br]a part of the ecosystem 0:17:30.198,0:17:35.692 rather than building everything [br]independently and everything separately. 0:17:36.661,0:17:38.721 And this is the current status. 0:17:39.624,0:17:42.668 By early results, we show that Wikidata... 0:17:44.945,0:17:47.075 a few games that we released 0:17:47.075,0:17:51.747 have triggered and proved activity [br]on the entities related 0:17:52.546,0:17:54.646 and a few follow up. 0:17:54.646,0:17:57.261 One thing that we have come up with, 0:17:57.261,0:17:59.971 as I have talked [br]to a few community members 0:17:59.971,0:18:02.388 is the PreCheck idea 0:18:02.388,0:18:09.088 that is basically providing [br]preliminary check about bulk uploads, 0:18:09.088,0:18:12.268 sampled preliminary check [br]by community member 0:18:12.268,0:18:14.478 and use that to generate a report, 0:18:14.478,0:18:16.185 make it easier for discussions 0:18:16.185,0:18:20.445 about whether this big block[br]of Wikidata datasets 0:18:20.445,0:18:25.095 should be included [br]or uploaded to wikidata.org 0:18:25.095,0:18:27.484 or it should be rechecked or fixed. 0:18:30.994,0:18:35.884 And there is another project [br]that is mostly a dataset project 0:18:35.884,0:18:37.300 called CatFacts. 0:18:37.572,0:18:42.642 CatFacts is datasets that we generate 0:18:42.642,0:18:45.552 about facts from categories, 0:18:45.552,0:18:50.231 the one that you see,[br]the Christian Scientist, just now 0:18:50.803,0:18:56.495 is actually an interesting outlier[br]of data points 0:18:56.495,0:18:58.344 from this effort. 0:18:58.344,0:19:01.861 This goal is to generate [br]the facts from category 0:19:01.861,0:19:07.363 which we think have been [br]very rich facts online that people... 0:19:07.731,0:19:10.087 that has been under leverage. 0:19:10.321,0:19:13.621 But before it can be fully leveraged 0:19:13.621,0:19:17.311 we need to make sure [br]that quality is good enough as well 0:19:17.311,0:19:22.261 and there is efforts [br]of putting it onto Wikidata Game 0:19:22.261,0:19:23.861 and there is effort that we're thinking 0:19:23.861,0:19:27.110 maybe building PreCheck [br]would help as well. 0:19:27.611,0:19:29.741 And it's still in early stage. 0:19:29.741,0:19:34.041 Feel free to come to talk us [br]about other efforts, 0:19:34.041,0:19:37.991 other ideas you think [br]about datasets we could provide. 0:19:38.499,0:19:41.539 The Bot, which is communication tools. 0:19:41.539,0:19:45.149 We know that Bot can do many things[br]like writhing Wikipedia article 0:19:45.149,0:19:49.841 but we promised[br]that we don't write actual article 0:19:49.841,0:19:52.597 but we mostly use it 0:19:52.911,0:19:58.329 as a way to communicate [br]from, let's say, user talk 0:19:58.817,0:20:04.397 to give us access [br]to large scale conversations 0:20:04.397,0:20:06.103 with the community members. 0:20:06.416,0:20:09.686 Explorer is going to show [br]all our datasets, 0:20:09.686,0:20:11.879 our toolings, their stats 0:20:11.879,0:20:15.491 and queries you can run on our things. 0:20:15.491,0:20:18.238 Stay tuned, this one is releasing soon. 0:20:18.960,0:20:20.933 And we have several other ideas 0:20:20.933,0:20:24.003 but I would jump [br]to this overall portfolio. 0:20:24.003,0:20:28.443 It would be several projects [br]to begin with datasets and tooling, 0:20:28.443,0:20:30.338 and what we are doing currently 0:20:30.338,0:20:33.190 is Explorer, Battlefield, [br]CatFacts and PageRank, 0:20:33.190,0:20:39.600 and there are some other upcoming ideas[br]like PreCheck, CitePool and Bubbles. 0:20:41.294,0:20:46.494 And this is one of the diagrams 0:20:46.494,0:20:48.574 that I want to show you. 0:20:48.994,0:20:53.385 We want to not only use [br]one individual project 0:20:53.385,0:20:54.734 to contribute the community 0:20:54.734,0:20:58.007 and also generate the training data [br]for the research, academia, 0:20:58.007,0:21:00.807 we also have an idea 0:21:00.807,0:21:04.519 that these projects may work together. 0:21:05.676,0:21:08.976 For example, the CitePool, [br]the system that we want to build 0:21:08.976,0:21:15.352 to allow people to easier find citations[br]for Wikipedia articles or Wikidata 0:21:15.887,0:21:19.316 but also use the Explorer [br]to display the result-- 0:21:19.499,0:21:23.079 it depends on the page rank [br]scorances of datasets 0:21:23.830,0:21:30.284 to determine how to rank the citation page[br]that we will recommend 0:21:30.423,0:21:35.630 and use the PreCheck [br]to do quality, sanity check 0:21:35.630,0:21:40.235 and maybe create [br]bulk batch reports by Bot 0:21:40.235,0:21:44.255 and the PreCheck will depend [br]on the Game as well. 0:21:50.727,0:21:52.566 If some of our community friends 0:21:52.566,0:21:55.476 have been following [br]the progress of WikiLoop, 0:21:55.476,0:21:59.005 we have been through ice-breaking phase, 0:21:59.655,0:22:02.335 we were trying to earn the community trust 0:22:02.335,0:22:06.152 because we know how cautious [br]we need to be 0:22:06.152,0:22:09.575 coming to contribute to a movement 0:22:09.575,0:22:14.704 that relies so much [br]on the neutrality and non-bias policies. 0:22:14.999,0:22:19.539 And we have gradually start to have ideas 0:22:19.539,0:22:22.545 about tools and datas[br]and find the direction 0:22:22.545,0:22:25.974 of how we can possibly [br]make this sustainable. 0:22:26.231,0:22:31.880 And we are looking into creating [br]long-term sustainability, 0:22:31.880,0:22:34.853 both internally and also externally, 0:22:35.160,0:22:38.654 both in terms of getting resource[br]and getting support, 0:22:39.024,0:22:44.545 also externally of getting engagement,[br]getting usage, and getting contributors, 0:22:45.568,0:22:48.122 starting from next quarter. 0:22:49.364,0:22:53.066 I want to quote Evan You, [br]who is a creator 0:22:53.066,0:22:58.588 of popular frontend framework Vue.js, 0:22:58.588,0:23:01.154 "Software development [br]gets tremendously harder 0:23:01.154,0:23:05.504 when you start to have to convince people[br]instead of just writing the code." 0:23:05.504,0:23:08.891 This applies to editing [br]Wikipedia or Wikidata. 0:23:08.891,0:23:13.261 It's very easy to click a button [br]and add individual articles 0:23:13.261,0:23:18.879 but also it's very hard [br]when you need to convince people. 0:23:23.330,0:23:27.440 I hope to leave some time for questions, 0:23:27.440,0:23:31.893 although we only have few,[br]probably one or two minutes. 0:23:33.229,0:23:35.993 Yes, so we have about two minutes. 0:23:35.993,0:23:39.085 So if people want to shout questions out, [br]I'll bring the mic over. 0:23:40.539,0:23:41.969 Hands up maybe. 0:23:45.433,0:23:50.273 (person 1) So where would I go to [br]at this moment if I would like to use this 0:23:50.273,0:23:53.563 to solve some of the problem [br]with chemicals, 0:23:53.563,0:23:56.553 where some Wikipedia pages [br]about chemicals, 0:23:56.553,0:23:59.663 they have a chem box [br]about a specific chemical 0:23:59.663,0:24:03.523 but are otherwise about [br]a class of chemicals. 0:24:03.523,0:24:05.746 Is that something [br]where WikiLoop could help? 0:24:07.750,0:24:12.923 I think that's the individual [br]domain expertize part, right? 0:24:12.923,0:24:15.523 If you are talking [br]about topics of articles 0:24:15.523,0:24:18.701 that are associated with specific topics. 0:24:18.701,0:24:21.131 We are trying to... [br]we might be able to help 0:24:21.131,0:24:26.301 but we are trying to tackle the problem [br]that is like more general currently. 0:24:26.301,0:24:32.531 And overall the goal is [br]to find the possibility of 0:24:35.201,0:24:39.354 empowering human beings productivity 0:24:39.354,0:24:42.204 and also trying to generate the knowledge 0:24:42.204,0:24:44.469 that potentially helps... 0:24:44.469,0:24:47.419 the training data that potentially [br]helps the algorithms. 0:24:49.682,0:24:52.231 (person 2) I think we have time [br]for a very quick one. 0:24:55.292,0:24:58.637 (person 3) Are you also going to do this [br]for search of data on Commons? 0:24:59.522,0:25:01.096 Yeah, we hope to... 0:25:01.096,0:25:05.239 If you are referring to Battlefield [br]or counter-vandalism tools, 0:25:06.451,0:25:11.615 yeah, we are planning [br]to expand it to other Wiki projects, 0:25:11.615,0:25:14.032 including Commons in Wikidata. 0:25:15.280,0:25:17.240 (person 2) I think that's all the questions [br]we have time for 0:25:17.240,0:25:19.800 but if you'd like to show [br]your appreciation for [Victor.] 0:25:19.802,0:25:20.932 Thank you. 0:25:20.932,0:25:24.612 (applause)