Welcome to the Social Innovation Lab. It’s co-created by hundreds of social innovators. We’re open until 11:00 PM every day. There’s a chef. There’s a kitchen. It’s a really co-creative space.

You don’t get hungry here.

No.

Oh, slides.

We didn’t know what you know about the organization we work in, so we have a quick introduction through the slides and then maybe something where we could explore about...

Is this pubic information?

Yeah. You could share it.

You can send us the PDF. We can publish after 10 days of editing our transcript.

Yes. I’ll ask Weiwei.

It’s good.

She’s into it.

She’s figuring that out.

OK. [laughs] I quickly talk and then see what resonates with you and go a little bit deeper. As a quick agenda, maybe the introduction of the group we work for, how digital we are, what we do here in Taiwan, maybe an interesting example from my own country -- I’m from the Netherlands -- and then some topic we could do to possible collaborations here in Taiwan.

The RELX Group, we help scientists do better science, very important lawyers win their cases. We help doctors treat their patients better and work for the insurance companies as well. There are four big areas. The one we work for is Elsevier. You might have heard of this. It’s a science publisher. You might have published yourself. I don’t know.

Yeah, I published.

Very good. We didn’t check your publishing record. Then there’s legal risk business or information for insurance companies. We’re one of the largest exhibitors. Without people knowing when they go to big exhibitions, often they’re done by our colleagues from Reed Exhibitions, like Comic-Con, for instance, if you have ever heard of that one.

Yeah.

A quick overview. We’re a large company or medium sized. €8 billion in revenues, 33 billion in capitalization. We’re 74 percent digital right now in revenue. More active in North America and then Asia and Europe. Lots of income through subscription model.

This is a bit how our digital transformation took place -- very happy about that because it’s been pretty fast for 2000. Digital was still a small percentage. Now it’s the vast majority. The orange is face-to-face. People still come to exhibitions in person.

Remained pretty much constant.

Exactly. That’s what you would expect, actually. The core drive for our industry is first a deep understanding of our customers, very high-quality content and data sets, good analytics, and also very powerful underlying technology. To go a little bit deeper, this is for the four businesses.

The four groups.

Exactly. We’ll just look at for Elsevier. We focus on scientists. We focus on the health professionals. We have amazing networks, 17,000 journal editors, almost a million reviewers. We have 10 million monthly unique visitors that come to our platform. It’s a great example of economy of scale there.

What kind of data sets do we have? We get 1.5 million article submissions every year. Of those, a third survive. Two thirds are rejected. The ones that survive, they’re usually rewritten two or three times. It’s quality selection and quality enhancement.

Today we have 15 million articles on our platforms, every single one peer reviewed. When you’re a computer scientist, you can trust them, just as I can trust, I guess the ACM or IEEE. We also have database for drugs, for chemicals, etc.

We do lots of analytics. We have got an example we’ll get into a little bit later specific for Taiwan. Today, we hired lots of technologist, so to say, invest a lot in technology as well. We have technology hubs. One is actually here in Singapore.

How many people in Singapore?

That’s a good question. We can double-check.

Hundreds or dozens?

Yeah, something like it.

I think it was more than...

No, it’s more than hundreds and thousands.

That’s good.

When I was there, I did...

It’s a regional hub.

Yeah, exactly. We’re actually building up on in Amsterdam, where I’m based, which is kind of funny because it’s our headquarters. From a technology perspective, we’re a little bit behind. They always had the big hubs in the US and in the UK.

Good to know. [laughs]

What do we do with the technology? These are the typical big data solutions. We have all these datasets. How can you make sense out of that? How can you focus on the quality? How can you do entity resolution, the link analysis, the clustering analysis, etc.?

If we talk about Elsevier, more the science publisher, today’s researcher has many, many difficult tasks to survive. Get grants, get published, do the teaching, attract the right people to work with, start collaborations, keep up-to-date with today’s research. I’m sure there are other things here, but the form size is too small for somebody of my age.

That’s OK. I can read them just fine.

Because you’re younger.

That’s right.

The approach from Elsevier is that we have the big data solutions, we have the rich content, and we have all these networks. If you combine that, it’s quite powerful.

You see on the next slide, very digital. 81 percent is digital today. Also strong on technology -- lots of machine learning, search, data visualization. One of the big databases we have is called Opus, which is all the digital metric data. Are you familiar with this?

Not very familiar, but I have friends working in that area.

Very good.

There’s a clear pipeline here, right?

Yes.

From the publishing, to the storage of data, to the applied knowledge.

Exactly. We are still a publisher, what we call a quality information provider. We still very much focus on getting the right content in. This diagram, this is where we see the different segments based on what is called the field-weighted citation impact.

Every discipline has a different way of working. A typical article in the life science will have 50 references, and therefore many more citations. In mathematics, you can write an article from first principles, nothing wrong with that. You can have one reference.

It’s far more difficult to get citations in mathematics, so you have to compensate for that. That’s why the field-weighted citation impact. It’s a proxy for impact, for quality. Here you see the orange is the Elsevier journals, and then...

The other companies.

...other large companies. Springer Nature is one of them and Wiley is one of them, Taylor Francis probably. You see that we do pretty well in the top 10 percent on that field-weighted citation impact. We do very well on the 10 to 25 percent. We do also pretty good 25 to 50 percent. We have very few journals in the lower...

That actually wasn’t the case -- I’ve been with Elsevier for a long time -- 10 years ago, 15 years ago. We really ramped up our quality. That’s something I think we can be proud of.

This is a bit of the transformation, very high level. Before we were a publisher and we just said, "Here it is. Go and read it." We invested very much like search for exactly the right information. Now, we want to give advice and do the right thing, answer questions.

When you’re a doctor, we don’t say go and read articles or search for articles. We will try to give you the answer you need to when you’re next to the patient.

Quite active also in managing research data. I don’t know if you know this approach. This is a bit on Nirvana, successful data management. Today, we are already lucky in certain disciplines if data is stored. Sometimes, it’s not even stored let alone preserved.

It’s accessible. They can discover it. You can cite it. You understand it. It has some kind of quality seal that is reviewed. You can reproduce it. It’s also that you can reuse it. This is a bit like where we want to work towards.

It depends on the field. In our own field, astrophysics, we do a pretty good job. Within chemistry, they’re still struggling at the bottom of the pyramid, so to say. We are developing and we have developed solutions for all these different parts. I’m not saying that we’ve solved everything, obviously, but certainly getting there.

In the government open data strategy, we’re working on the same thing.

Exactly.

Taiwan has been, for two consecutive years, ranked the top country, international, on open data. I think that’s because we have a very firm commitment that everything that is freedom-of-information accessible, it must be provided in a machine-readable manner. It’s not just for the right to know, but actually the right to reuse.

And the right to search.

Exactly. This doctrine is essentially saying machines are people, too. Just like we need to provide blind people with universal access, we need to provide AI with equal access. The jump from comprehensible to reusable is a strictly followed guideline in all the levels in Taiwan government.

That’s great.

It’s a part of the pyramid.

Excellent. You see also that it’s used a lot. People find it and they download it and cite it.

It increases trust. People, like in climate research or in air quality, if their data are not compatible, then they are not actually doing science because you cannot compare the different models, which is why we all follow the international standards in such things.

Some of the things we do specific for the datasets. It’s linking from articles to the datasets and repositories. We have our own data repository.

I see we use the same open data license, which makes things very easy.

Exactly. We’re active in most of these more technical communities. We want to stay with the right standards as well, and also contribute to developing some of the standards.

In Elsevier in Taiwan, how many people work here?

In the realm of 30 people.

Like a branch?

Exactly. We see here on the bottom, we provide the right content, journals and books, a drug database. We also have the right platforms. We already mention Scopus. There’s been Mendeley. It’s a bit of Facebook. You probably have heard of Mendeley. SSRN, this is for the pre-print, which started in the social sciences. We’re now doing it in other areas as well.

Here, to provide this insight, we have the more analytical solutions. You see a little bit like an evolution going from content to platforms to insights.

What’s Engineering Village? This one is new to me.

It’s a specific collection of data, which start mostly around publications and the citation database for engineers.

I see. What kind of engineer, like civil engineering and so forth?

All engineering.

Software engineering, everything?

Yeah. It goes back actually a century, so there’s lots of information there. Knovel is also something in the engineering space which we acquired.

I already mentioned Scopus. It extract the information from 5,000 publishers, certainly not only Elsevier. Then we have this analytic solution, SciVal, which is based on Scopus.

I’ll skip all the Chinese, but I’m sure it makes sense to you.

It makes sense.

Here, we provide the data for two rankings. Take, for instance, Times Higher Education Ranking. They have this weighted impact of teaching, research, citations, international outlook, and industry outcome. In that, big chunks are provided by Scopus.

Look at the QS World University Ranking. We provide all the citations, so it’s 20 percent. They also use the Scopus database to find the right people to write for the academic reputation. This is kind of a questionnaire. Weiwei’s going to tell you about the very specific ranking we support here in Taiwan.

We just recently partnered with Global Views...

GVM.

Yeah, GVM. They have special list for Taiwan university ranking.

You make have to speak a little bit louder. Sorry.

This is a special project for Taiwan university ranking. Here you can see five different measurements. For the research component, data comes from Scopus and the SciVal metrics. I think this is in July this year issue. Based on the metrics that we discuss with them, we came out like...

You can see on the website. They have four main categories for Taiwanese ranking. One’s for comprehensive university, public university, and also medical university. Different categories use different...

Metrics.

Metrics, yes. The 18 university president came to receive the award. There is a press conference that we hosted. There was a...

Launch event.

...award presentation.

You help with four of the five indicators?

Yes.

Yep. We don’t do teaching performance. We don’t have that information.

They also add their own measurement. They add research center and do their own metrics. They combine with ours to come out this ranking.

Societal impact, this is media mentions, right?

Yes.

It’s a bit of a proxy. [laughs]

It’s totally a proxy, but I’m not going to challenge that. What I’m going to say is that what kind of media links are you doing for so-called online media? Do you mean online media as in other online versions of traditional media, social media, or blogs?

I think it’s both.

Yeah, it’s both.

Certainly, it’s the social media as well.

There’s a social media part analytic in your database?

It’s called Plum Analytics.

Traditional media, too, like CNN. Any news or any exposure on CNN is traditional news online. This data state also got collected and stored in our database.

I mean more of the independent blogs...

That’s a good one.

...things like Twitter accounts with a lot of followers, but they are not traditional media.

No, but certainly the social media is there, LinkedIn, Twitter, Facebook, etc. I don’t know all the blogs. That might be more difficult to get a good overview.

It’s a different world.

I can double-check.

In humanities, like philosophy and so on, they rely heavily on self-organized forums for societal impact.

Exactly, and that’s where most of the discussion takes place.

Exactly. For example, in mathematics, maybe in Terence Tao’s blog...

...that’s where the conversation happens. It’s, of course, a proxy, but I’m happy to see that you’re including various different forms of social media and online media.

We’re actually thinking very hard to develop better metrics, also because we get this demand from the universities and the funders. An easy one is economic impact. We can make this link between articles and patents.

That’s the corporate collaboration part.

No, actually those are articles which are...

It counts patents.

Yeah, but here the one is mostly the articles which are published by, say, a researcher at a university and a researcher in industry.

Yeah.

That’s academic-corporate collaboration. That’s are easy. For instance, it would also be very interesting for you do research and at some point -- say, in environmental studies -- you shape a new environmental policy.

Policy impact.

It’s reflected in the legislation. We’re trying to explore that, but it’s very tricky to really follow that path in a way and then really link policy with research.

It would help, of course, if at the end of our policy-making process, we link back to the citation.

Absolutely.

That would help you, right?

Yeah.

[laughs] We have our own government research bulletin board. I’m sure you know about our website. Everything that is publicly funded, we’re now having a policy of having the policymaking context linking back to the taxpayer-paid research. That provides you with a clear back-link to the research that’s done. If that happens to be also published in your database, that’s a very clear link.

We’re using the Sustainable Goals as the primary index.

Sustainable Development Goals, SDGs?

Yeah, the SDGs...

Perfect.

Which is what I’m wearing.

...as the main index, especially for the environmental and social impact, exactly as you said. There’s very few reliable, internationally useful indicators in this area, but SDG at least everybody has agreed to. [laughs] It’s a proxy. I would also admit that, but it’s a good start. [laughs]

It’s a start, and this implies that your policies are based on scientific evidence. [laughs]

Yeah.

That’s not always the case in all countries, right? [laughs]

I’ll refrain from commenting.

Very good. How is Taiwan doing? We have all the data compared to other countries. Taiwan is in the top 20. It’s number 20.

We’re the global average in the sense of field-weighted citation impact. [laughs]

Exactly. You’re very, very close. Here, we have another graph where you see an interesting correlation. Here again, you see the field-weighted citation impact versus international collaboration.

Countries like Switzerland and my own country the Netherlands, we have loads of international collaboration, of course. We are a bit smaller, so that also forces you. Taiwan is not that huge, either. You see that they do very well within the impact as well.

One advice that I would also give when I meet a university president, I say, "Can you stimulate international collaboration?" It seems to be really...

Yeah, that it’s a worthwhile goal.

Yeah.

We’re slowly rolling out a bilingual policy where in kindergarten we start, maybe next year, to teach immersive English. By the time they’re grad students, maybe we switch to add English as one of our official languages.

For master’s level?

Yeah, but his whole process will take a decade or so. Slowly, we’re rolling out English as one...

You think the language is really the bottleneck?

It is one of the bottleneck. Language acquisition is getting easier with machine translation anyway. This is more of a familiarity of the field. Instead of relying on people to translate, we can now use machine learning and basically jump in into any discussion.

Even though there’s bound to be some errors in machine translation, at least that gets people to know the right people on the international communities. Not afraid of using broken English, because most people are speaking English as the second language anyway...

Not our mother tongues, either.

That’s right. Not afraid of publishing in a not-perfect English is one of the psychological barriers that we have to get across.

Absolutely, but a bit also cultural factors. I lived in Japan and I go there a lot. Actually, the number of Japanese scientists willing to go abroad is going down.

I see that.

Here, we compare Taiwan to other countries in output. Taiwan is stable. It’s the green one with the green arrow. The one that really jumps out here is, of course, China. Here you see the international collaboration. Of course, Hong Kong and Singapore do very well there. All the countries are more or less grouped together. A positive is that Taiwan is going up a bit.

You see the same kind of clustering for the impact. Again, Singapore and Hong Kong are together, and then all the other countries are more or less at the same level. The only thing that really changed over time is China. That was far below the world average already in 2008 and is now on par with the world, Taiwan, South Korea, and Japan.

The more people who do science, the greater the total contribution.

Yeah, but this is a quality measurement.

I know.

You often see it...

It’s a good thing.

Yeah. We have lots of detailed data. This is at the university level, just to say that we have that as well if you’re interested. If you want to look at your favorite university, we have that.

I’m a junior-high dropout.

I don’t have a favorite university.

Sorry, we don’t have your junior high in here.

We also mapped out Taiwan’s New Southbound Policy. Maybe Weiwei wants to say something about this. I see something upside down.

I think there’s a special report here.

There’s more analysis in this one. These are aligned with the government policy, using the government’s data to pull out the Southbound countries. Majorly used Southeast Asia. You don’t see Australia and New Zealand here. Usually, we recruit students and new professors from Southeast Asia countries, so that’s why I focused on these countries.

Also, India, of course. There’s a lot of India collaboration.

You can see India. You can see a lot, in terms of collaboration, Taiwan and India’s. You see India is top one or two.

I’m well aware of that.

This one is collaboration, publication in these five different areas. The first one you see, in engineering, the country that we collaborate with is India. The second one is in Singapore. You can see different country in different subject areas.

This one, as you can see, it’s only measured Taiwan research performance in 2012 to ’16. You can see in this I only placed the top four subject areas. You can see engineering is our major focus, and then the second one is medicine, and computer science, physics, and astronomy.

This chart is different from this. This, you can see what most of the collaboration comes from, which country in different subject area. This, you can see...

The country.

Yeah, by country, you can see. With India, the highest collaboration is physics and astrology.

Astronomy.

Yeah, astronomy. Sorry. [laughs]

I’m sure they’re very good at astrology as well.

Yeah, but they don’t publish much about it.

You can see the...

My background’s in astronomy, so I’m sorry, I...

I know.

Before we talk about your astrological sign.

You can see by the number of publications and also the quality of papers.

This is really insightful and very well-prepared. Is there a reason why you choose a lighter color over here? It’s just because it’s less collaboration?

No, this is just the design. I send the data to our...

Designers.

...the backend design, yeah. This is our...

This is more like your color palette.

[laughs] That’s right.

It’s very easy on the eyes.

[laughs] I thought I would bring a Dutch example. We worked with the Dutch government. Essentially, they say, "We want to live in this knowledge economy and to enter our societal challenges, for instance, around affordable food or sustainable food production." The Netherlands is the number one food producer in the world producer in the world after the US, just as an example.

It’s this mix of technology and innovation policies. You see here how the Netherlands is doing compared to the other European countries. Again, pretty good in quality. Denmark is a little higher, unfortunately. Just in the middle as output, because we’re a mid-sized country or small, whatever you want to look at it. We cannot compete with Germany.

We looked at specific top technologies and we really zoomed in to very detailed ones. What jumped out is that the Netherlands does pretty well in the life sciences, for instance, stem cell technologies.

Genomics.

Exactly. What also jumped out is, for instance, microreactors. This is more chemical technology. Imaging technologies and optical, mechanical engineering solutions.

Then we see, "Who exactly is doing that research?" You see an example here for nanotechnology. This is at the technical university of Delft and Twente. This you see the number of publications and, again, the impacts and who would be their competitors abroad. MIT, CIT, and CRS and what is their quality?

What will jump out in this whole list is this ASML, which is a company, very big in the Netherlands, but also very active here in Taiwan. I think there are thousands of people working for ASML here in Taiwan as well.

That’s a bit. Overall, the conclusion of this study is very research-intensive. The Netherlands really perform way about our size. Very good in all these key technologies. We outperform many other European countries. I mentioned already it’s very much in the life sciences, chemical, and engineering.

Very international, more international than others, except for the Swiss, but they’re not in the EU. Then, there’s a very nice mix of mature technologies and new technologies. This is all based on data. This not interviews or expert opinions, this all comes from our database. This is a bit of an introduction.

This is great. Like almost pushing the limit of what two-dimensional media can do.

Exactly.

I’m sure you’ve used virtual reality. There’s more dimensions to work with.

Yes, absolutely.

Thank you for sharing. This is very exciting work, and certainly, as we already mentioned, in Taiwan we are looking towards expanding more international collaborations. As a minister in charge of social innovation one of our goal for the next four years with the social innovation action plan is to develop a way to measure social impact and environmental impact of academic activities.

In Taiwan, we have a plan called the University Social Responsibility or USR, which is like CSR but better...

...because it can make the students complete their custom projects just by working and focusing on a community issue. Just as you said, maybe they work on biodiversity, maybe they work on ecological economics -- not astrological astronomy.

Ecology and economy, and to figure out the way for those two fields to bridge together, maybe as a social enterprise or some other way to create value together. If it doesn’t work, it’s science, so people learn from what doesn’t work, because they’re still in the university. Their parents doesn’t feel like is it a failed entrepreneur, [laughs] so maybe it does work.

Then there is academic contribution, so their professor will publish, but then we face a problem of how to define the social and environmental changes. That’s not necessarily policy.

It could be a change in people’s mindset toward endangered species, toward recyclable, renewable materials of viewing plastic not as a waste but as a way to regenerate biofuel or whatever, so it changes a social dynamic, but how do we actually measure that?

Most of the ways we learn from the EU especially is either a SROI, which measures the social return of investment in dollar terms, like how much time or money down the line are we saving by intervening before the social problem happen. That’s one way.

So a preemptive thought.

Preemptive one. The other one of course is like the...I don’t know whether you know the B Corp Movement, the B Lab Movement that declares a corporation as a benefit corporations meaning that...

The B is benefit?

Yeah, yeah. The corporate declare for themselves what kind of social environmental and governance impact they have like triple bottom line, in addition to the original bottom line...

Financial, yeah.

The financial ones. This is more of a huge checklist of hundreds of micro items, and they assume that if the company is doing a majority of this well, it’s a well-balanced company that creates certain impact, but as you can see, this is a proxy.

It doesn’t actually measure the actual impact, but actually it measures the shape of governance that could lead theoretically consistently to positive impact. Between those two paradigms, I’m sure that we can develop something that is specific to university and graduate level studies that works on the basis os sustainable goals, but doesn’t end with sustainable goals.

Because the goals just say, "You know, by 2030 we need to be here." They don’t say how are we getting there and whether doing this movement is actually moving forward or backward. It doesn’t say anything about it. [laughs]

I think in the next four years there’s going to be a lot of funding and a lot of academic interest in bridging the corporate ESG accounting and the social return of investment and of dollar conversion into something that can measure the societal impact in a more consistent way that also work across countries.

That is an active area of research study. If you have any insight or anyone you know who has any insight will be totally thrilled to have this conversation.

We’re also active in the area of the SDGs. We have a SDGs resource center. I’ll send you the link and see if there’s anything in there...

Yeah, we’re integrating that into our curriculum, so anything that helps people to raise awareness is a good thing. We work with international organizations already. There is a GSMA, the association that does the GSM, and they are mobile telco providers and so on.

They are actually shaping their SDGs Explorer like this, so you can very clearly see that we’re in a world by engaging which telco operator, which SDG goals are they making.

We’re going to do the same mapping for social entrepreneurs for USI and CSR here in Taiwan, so we’ll have consistent data maybe down to the target level data for everybody to do data science around. It makes it easier for social entrepreneurs to find the academic support that they will need, because then we’re working on the same goals and targets.

Of course, we should work together. If you have a resource center that is somewhat similar to this shape, I’m sure that we can collaborate more closely.

Perfect. What is this, say, for Taiwan?

There’s quite a few cases for Taiwan telcos, usually the Digital Opportunity Center, and then there’s also UA4G project which makes use of ET. Japan uses the natural disaster recovery which we are doing a compatible called the CI, the Civil IoT project, which you may be interested, because it is the first time that we’re using citizen science data alongside academic science data.

Usually they’re two different worlds, but now we’re making sure that people here are using advanced data science. They can upload their own air quality measurement data, snapshot it, and store on a distributed ledger to make sure that we don’t modify anyone’s data.

In any case, making sure that people can see the whole air quality through citizen IoT, also our law enforcement-based IoT and sensors.

And it’s measured to smart phones, right?

That is exactly right. There’s citizen science, what we call AirBoxes that are all over Taiwan, and we integrate in a way that ensures their quality.

For places where the citizen scientist cannot go to, for example, I usually use this map where there’s citizen science points that measures the air quality and correlates to the academic research on these things. It shows the digital gap in Taiwan, I guess. [laughs]

These are places where citizen scientists are active, and the universities and environmental protection agency...

And the color means the quality

The color means air quality, yes, PM2.5 in this case, and of course we flew in the places where the civil society cannot go to like high mountains, and even in the middle of the Taiwan street, because we’re going to have off-shore wind turbines...

...where we’re going to put air quality permanent sensors on all the turbines to help getting the root cause of which air pollution comes from out there or from inside of Taiwan. The important thing is that all this is open innovation meaning that it’s open source.

The data is out in the open so that everybody in the word, even though it’s Taiwan-innovated, they can download and run it locally in Asia and build on this, which in this case informs the Japanese community through code for Japan, so this is a cross-country collaboration right there.

In addition to the Taiwan cases which I talked about, we’re also working with NZ to make sure that our civil IoT for water pipeline, like pressure and flow and so on, can be used to detect water leakage using machine learning.

They didn’t use to have a water shortage problem, but now because of climate change they do, and so they’re partnering with us for three months to codevelop a AI algorithm that works for both countries, and I think this is a new kind of diplomacy. We call this warm power. Not just sharp power, warm power, in the sense that people can share easily.

It’s the old science diplomacy.

Right, exactly. The real issues that we’re facing with, and making sure that, for example, Italy, just imported our model of budget visualization. This is a Taiwanese round, but they are setting their Italy chapter to do budget visualization.

Controversial budgeting it.

Yeah, and analysis and a real-time conversation. What I’m getting into is that this is by Nature International, because of open data, open source, and open science, and any work that you...

And open access.

And open access. Any work that you have in terms of the environmental data, we’re happy to integrate into the CI system. Which stands not only for civil IoT, but also collective intelligence.

Oh, OK. Good, we’re happy to share what we have. On a more mundane level, the publishing industry is very active also around SDGs. They’re going to launch an SDG book club for children in all the six UN official languages.

For every month, so for 17 months in a row, there will be books in Spanish, for kids in Peru.

That’s great.

Around say, water, or about gender issues in the Arab world for a child in that...

I wrote some system that can take simplified Chinese and automatically produce traditional Chinese.

Oh, OK, all right. The children won’t notice when they read it.

It is one of my early contributions in year 2000 or so. I think BBC and a lot of media use it. If you have those kind of books in simplified Chinese, and you’re willing to contribute to traditional Chinese books, the translation is entirely automatic nowadays. You don’t have to hire a translator.

I hope it will all be digital, by the way. That’s the assumption here.

Yeah. If it’s paper based, then we’ll have to scan it, anyway.

Exactly. Part of the publishing industry is extremely advanced. You’ve seen Elsevier or Relics. It’s all digital. Till and book publishers are quite conservative.

In Taiwan, broadband is a human right, and we make sure that all the schools have access to broadband access, which is part of our education strategy. Because of that, anything that you publish digitally, you can be assured that the indigenous people, the remote people up in very rural areas, they have equal access to them, including tablets, and so on.

Yes. Very nice. Anything else from your side, Weiwei, you want to bring up?

I’m interested to know if any projects that you’re involved in is about to collaborate with the Ministry of Science Technology’s project?

I’m on the weekly Board of Science and Technology meeting, although I’m not the BoST minister. That’s Minister Wu-Jen Cheng. When it comes to digital or technological issues, like the civil IoT, then I personally review the project and lead the technical part of it.

The CI project is actually led by the MoST, although it uses data from the EPA, from the MOEA, from all the different ministries. I think one of our contribution is that we make sure that for such cross-ministerial and cross-municipal projects, we always use a domain name like CI.taiwain.gov.tw.

Our AI strategy is AI Taiwan. Social innovation is SI Taiwan, and our digital strategy being smart Taiwan. Instead of you google and find five ministries I’m sure is very different and confusing message, we make sure that it’s just taiwan.gov.tw now.

The main work that I’m involved with is AI Taiwan, CI Taiwan, SI, for social innovation, Taiwan. All the three have Ministry of Science and Technology involvement.

Last time that I was here, I met with Minister Chen.

Yeah, he’s great.

And smart.

Very, very smart.

It was a very good, good dialogue. He came well-prepared as well, with his team of directors-general.

For AI Taiwan, for example, there’s a grand challenge going on to work on speech recognition, and developing Taiwan-based AI speech recognition that includes semantic understanding and so on.

The next on is going to be IoT cyber security, I think. There’s going to be quite a few sandbox -- that is to say, limited time, breaking the law -- trials on automated vehicles as well, and the fintech sandbox as well.

All of these, although not directly MoST-led, there will be a MoST component in it.

Good, exciting.

Any informational data analysis provided by the STPI?

Yeah, sure, and the NAR Labs and so on.

We also work with them.

I am aware of that.

For not just MoST, MOE, for the research performance projects.

What about NARLabs, ITRI, III, as well as the other think tanks?

Yes we have worked with NARLabs, STPI for some projects such as University rankings and researcher mobility but haven’t worked with ITRI and III yet.

OK. Traditionally they work with e.g. the Smart Electronics Industry Project Promotion Office.

Yeah, I think ITRI and III probably just focus on how they can directly convert the science work to more commercialization, right? [laughs]

That’s true. What we’re trying to do now with the SDGs, though, is to co-develop a Horizon Project, just like the European Horizon Project, or the Japan Society 5.0. We’re trying to say, "By 2030, we’ll be here, but what about 2050?"

How to get there, and what’s our collective imagination. The think tanks are now tasked with not just applied science, which is always with a horizon with only five years or less, but actually the envisioning of the society.

There’s that part, more like speculative design part, and also a social conversation consultation as well. That actually, you need a lot of scientific, maybe more scientific research and data-backed, because that’s how we to tell the science fiction from science nonfiction, right?

As far as I understand, many ministries are working together all toward this common horizon-setting scenario planning. This is a new development, like new as in this year.

Excellent. Any questions from your side?

No, I think this is a great contribution. It’s good to know that in Taiwan, your branch is doing well. If there’s anything that I can help, especially in terms of SDG. That is my main mandate now, is to make sure that our voluntary national review include not only the governmental contributions.

In Taiwan, frankly speaking, the civil society space is expanding, whereas everybody near Taiwan has a shrinking civil society space. We are the most equipped to do social innovation, but it’s not always accounted for. It is not always linked back to research.

The professors, they do a lot of work, but they don’t always reflect that in their scoring.

We really need to build a system where their work in creating social impact...

Get the credit for that.

...about getting the credit. The same as you have one of the pillar of the excellence in teaching. We also want to have a pillar in excellence in creating social change. That is more than media exposure, of course.

Which is on the agenda in many countries, but difficult to solve.

It is, which is why we need international collaboration.

In the UK, for the REF, the Reinsurance Evaluation Framework, they used anecdotes, essentially.

I know.

It’s difficult to quantify.

I visited the Ministry of DCMS — well, the D and C and M, I have not yet visited the S — they are experimenting with different ways to measure. Many of those are anecdotal at the moment. They are working, of course, with Nesta and folks to try to develop new methodologies.

As far as I know, they all of this are all in early research phases. It’s the same for all of us.

Good, great. Thank you for your time. I really enjoyed the conversation. We’ll send you a...

The PDF, the digital counterpart?

A PowerPoint, or a PDF, yes, absolutely.

We’ll publish alongside the transcript, which we can edit together for 10 days.