Ep 7. Data Transparency and COVID-19 (MY) Part 3/3

SUMMARY KEYWORDS

data, information, contact tracing, pandemic, organization, ministry, testing, moh, reveal, malaysia, underlying condition, cases, government, death, person, technology, managed, tracing, exacerbated, release

00:00

This is the final installment on the COVID-19 series on what's happening in Malaysia, we speak to idealer Raza from kini labs.

00:35

Welcome to the security law podcast

00:45

your opinion, what additional data that the government can and should provide to enable people that you meet prof to actually look at the data and present a view where it's much more concise, gives us enough depth without revealing personal information, and giving us a chance to look at this whole pandemic. from a data perspective, rather than just a holistic number,

01:18

I would say that there are a few low hanging fruits data that they already have on hand, which is which they already have cleaned and sorted for their internal use, which they can share. One is testing figures. I don't understand why Malaysians are not provided testing figures every day. But researchers in Oxford University get them because we have that data from that page, our world in data, which is run by Oxford researchers that can be provided instantly. So because we know that the more you test, the more cases you find. And then if we reach a certain threshold, I say now we're at 5% Positive positivity rate, it means that that there is more in the in the community that we are not testing enough, for example. So those those that kind of basic information is very easily shared can be shared. And it's already available for within KVM and already sorted for that internal use, they can release that very easily.

02:29

I'm not going to find it really funny that someone is going to get Malaysian data from Oxford and not from Malaysia itself. I'm not sure what to say about this.

02:38

Yeah, it's very infuriating, is that right now, we have to go look at that website. Hello, world data, which is run by Oxford to find testing figures formulation.

02:51

Believe Malaysia also has an initiative on open data, where they look at publishing data, I've had some interesting looks into some of the data. Like for example, there was one data set that I saw, which was on number of complaints a particular ministry had about technology that they use desktops, laptops, printers. So I'm curious as to why they wouldn't use this particular channel to publish that data. I mean, it should be the same initiative within the same government

03:21

by right but things are people around the ministries to kind of move in silos. So they may not think that they have to work with this other agency or this ministry that handles the data, for example. But even that initiative, I think, has lots of holes. Like it's just it's really up to the agency or the government ministry to declare what data they want to send to that organization. They're not obliged or compelled to, to reveal anything. In the I mean, back to your first question of what information should be available, when is testing data? And then the other one is the locations I said, which they already have an easily published, like how throw us and also to use my sujeto to inform people instead of having contact traces, call these people one by one. And thirdly, I think there needs to be a revision, or how they release data on at McKim level or locality level. They need to understand that people don't know what their motives are, they know what their Tommen is. Maybe they're like, a bank or whatever, but they don't know what they're mocking is. So it's kind of an effort in futility.

04:44

I completely agree. We got confused whether it was in bottling Damansara or Shah Alam and I just realized that Charlene was actually not even a Mukim. So it's a confusing thing that's going on right now.

04:55

I don't know where I am right now. Thanks very much. When we talk about or when we barring about data trends, digital transformation, whereby the organization is expected to, you know, take on a lot of modern technologies and in do wonders, create new revenue streams and things like that. Usually, the first thing they will look at is the data because data is like a goldmine. Data is the new oil data is the new valuable thing. And what a lot of the vendors the technology vendors will say is, if they're there, if they exist, silos, data silos in your organization, the first thing you can do is to unite all the data, put it all into one central place, so that you can do more with it. You can do analytics on it. You can like glean insights from it, you can predict stuff, and so on and so forth.

05:56

I feel like there is an effort to do that, through the data.gov.my website, is run by one, I believe. So the idea is for them to have all the data you need in from all the various agencies and ministries. But ultimately, because there is no legislation compelling the ministries or the government to reveal the information. It really depends on who wants to share the information? How

06:33

do we look at the government sharing more data? Or what what do we what can we do to help stem this pandemic? Is

06:44

Molly the person with all the answers for this? I mean, everyone's grappling over how do we get these cases down? But because of the area of work that I do, I feel like those initially, those three things that I suggested should be made public, the locations, sought out the Mukim thing in and reveal locations according to time I know. localities that or even your street, not anyway, find a better way to reveal information, you can just cut it, and also provide us with testing data. So that's the three things that I think can be done very easily and quickly. But whether or not this stems, the pandemic, I mean, I don't know, I don't know whether people are actually going to look at this and realize that, okay, we need to change, change the way we behave, and start, try trying to trying to limit our movements or whatever, you know, so I'm not sure whether that will actually lead to reduce independently, I mean, controlling dependency. But I agree with you earlier that if you have quite a lot of the candlelit data out there, there may be more people who will be able to pitch in to do more analysis even provide solutions technology solutions to contact tracing. And that will definitely help in coping more infections.

08:17

Like what you hear so far, make sure you never miss a show by clicking the subscribe button now. This podcast is made possible by listeners like you. Thank you for your support. Now, back to the show.

08:34

Besides contact renting, do you think that there are any other technologies that we can use? I mean, Singapore's introduced proximity tracing. There's also the other flip side where there are countries that don't use contact tracing, but have had success managing this pandemic? I don't think there

08:53

is no, there's no country that doesn't use contact tracing, contact tracing is the most basic in reality, so like controlling infection. That's the most basic thing. They may not use very advanced technology for contact tracing, they may do it very manually by interviewing people and going door knocking or calling that still contact tracing. Everyone does contact tracing is just how you do it, and what sort of advanced technologies you use. So when you say like, how else can we do it? Besides contact tracing? That's no contact tracing is what everyone does.

09:29

But actually, I was just having when I was just reading this article some time ago, and something that you mentioned also was that that the data is pretty much sitting between different government organizations, right. So somewhere in April of 2020, it was mentioned that the data sets are that basically it's a collaboration between NSA MOH Mompou MCMC. And then in November, same year, it was said that the data is monitored by Naxa and NSE. As a common person It kind of gets to me that, you know, I kind of tend to think where exactly is my data is if the one organization is across multiple organizations, who actually has visibility across it all? Is it managed separately? Because you know, it kind of the most data sources you have, from a security standpoint, I think it becomes more challenging to manage that. So I'm just curious about it. It's kind of an open ended question not necessarily looking for an answer at this point.

10:31

I, myself am also curious about where my data is sitting, you know, on the Meisinger. website, I read a visited it, which was quite some time ago, a they have like information, or in the FAQ section, they will say that data is held by I can't recall which organization, but it sounded as though only MOH personnel would have access to the data, meaning they will be able to view the data, if I remember correctly, also, they will be holding the data for a certain amount of time only after which the data will be destroyed. But I think is, is causing a lot of questions as well among my Siddhartha users, or those who are thinking to join or to register to message or to app whether this is actually followed through. Do you have you heard anything about this?

11:21

This is went through the FAQ. Now this is what it says it says that the information is managed by key him with assistance from M key and NSC or a man who assistance with assistance by NSE and Mongo,

11:41

the only person we should ask or we can ask is either mostly or kick him? Well, technically, they should have access to data. But I also think that many other people may have access to this, we won't know until someone actually does an audit and presents a report and says, Yeah, this is how we verify this system.

11:57

The wording is interesting, because when I first read it is it sounded as though as few parties as possible would be viewing the data managing it. I don't know, does it? Does it involve being able to see the data? Yeah, so is the wording of the FAQ is interesting. It does not reveal much, it's very obscure, still pretty obscure. And as much as possible, when it comes to something like this is like a personal information, you know, usually the ideal situation is that as few people as possible,

12:32

actually, the reason given by KKM, for not sharing information with this task force, or COVID, is that they are not part of KKM. Therefore, they do not have permission to access that information.

12:51

But there are still many questions that remain like, for example, the one that we just found out about my structure and how data is being managed. How is it being handled? And what actually happens to the data? If and when the pandemic goes away? There's no no clear nuclear answer about this. But I think I can,

13:13

I guess the only people who can really answer it will be the people, the moisture people, maybe you should get them on the show. Yes,

13:20

this is very good. Yeah.

13:23

responsive. They're quite responsive. In the email, I noticed some mistakes in their data. And I to them, and they call back to me, and so get

13:34

it out. Oh, awesome. Maybe we can get them to be on the show.

13:37

Yeah. How do you think we should?

13:40

What kind of data is missing as a data scientist? What's What do you think should be released for you to be able to get a better idea of what's happening in relation them? So COVID-19?

13:51

Actually, the overarching question I had for you earlier was do you think better data could equip us to deal with a pandemic? Because to me, I always rely on the fundamental theory that what's not there, we could not find it. Okay, but in this case, I do agree if we have more information, very granular level data, such as whatever you have mentioned, I do agree in terms of locality and stuff like that. You mentioned testing figures and all those they will help. But what I needed was the causal relationship data is not sufficient. For instance, everyday we know about death, the number of deaths are being reported. So what I want to know would be, are they death numbers because of COVID? Or is there an underlying condition? So it is these type of informations that are hidden from us that makes it very hard for us to draw conclusive evidence of what we're reading and meaningful information from the data, which to this date. It still depends on the government and what they want to release, I'm not

15:01

sure the underlying conditions are released

15:04

already. But the thing is, in, for example, when we have a person who's declared, you know, clinical leader, then what is the cause of death written there? Do we actually know that the person died because of COVID? Or is it that the person died with COVID? Is there?

15:26

Yeah, so a few few months ago, I don't remember exactly when the Ministry of Health decided that they wouldn't be putting in cases where someone is was positive for COVID, but died for other reasons, for example, that they died in a car accident, they would have included in the list. But in cases where they don't know where whether COVID has exacerbated the other illnesses, and led to death, then they will include it in the deck list.

15:57

So in that sense, and I will say that the numbers that are reported, would, I would like to put it, quote, unquote, misleading, because it creates different interpretation different way where we would decide what next what needs to be done in order to cut the issue.

16:14

Anyway, so because they are excluding the numbers of people who died in car accidents or suicide or whatever poor was difficult. So the only deaths that are included in the COVID deaths lists are debts where it cannot be determined whether it was due to COVID or not, or it can be sure that it is COVID. And I also think that it's difficult for them to say whether it's an absolutely because of COVID or not, isn't it? Because we don't know some of them died of a stroke. And that could be exacerbated by COVID or not, we don't really know that for sure.

16:57

So on that and then I will say that the challenge then is how do we draw meaningful correlations from the data which we have or in this case, which we do not have, because we are not, we are not able to get it,

17:09

the data is available, but whether whether the correlations are confirmed or not. I think that scientists all over the world are trying to work that out.

17:26

Thanks for joining us this week on security lab. Make sure to visit our website at security lab dot Asia, where you can subscribe to the show in iTunes, Spotify, or via RSS, so you'll never miss a show.