Marrying Data Science and My Love for Travelling: A Well-spent Summer Getaway
In the European academic calendar, summer is the longest holiday: a great time to take a break from classes and explore a different part of the world. I am currently pursuing a master’s degree in engineering at Telecom SudParis in France with a focus on artificial intelligence, and I sought an opportunity that would allow me to gain some professional experience and learn about a new culture. I came across several options in lucrative institutions such as banks and private tech companies in my search. However, I was more inclined to work for an organisation in which my work would have some kind of social impact as I’ve spent most of my life living in various developing African and Latin American countries.
United Nations Global Pulse stood out to me and its mission to accelerate the discovery of big data innovation for sustainable development and humanitarian action seemed like a match with my personal ambition. I have always loved travelling and getting out of my comfort zone, and the Lab in Jakarta seemed like it would be a good fit. I knew very little about Indonesia but just enough about data science, so I embarked on the journey with no doubt this whole internship experience would be a fair exchange.
I guess it wouldn’t be a bad idea to make this into a travel blog and talk about Indonesia’s breathtaking landscape, the many cuisines I tried and the amazing friendships I’ve made, but I’ve decided to focus more on what actually made the internship itself such an enriching experience for me.
These are a few things I took a mental note of along the way:
- Machine learning and Artificial Intelligence can be applied in almost every field: from medical diagnosis to autonomous vehicles, these methods have and are changing our daily lives. However, as future data scientists, I can say that most of my classmates are more attracted to finance and consulting jobs than employment in the development sector. Certainly the salary has something to do with it, but I also think that as students, we are not aware enough about data science applications in the development sector.
At Pulse Lab Jakarta, I had the opportunity to work closely with the data science team and contribute to several projects that involved applying machine learning to inform government decision making. From exploring how to predict wealth using mobile phone data, creating a natural language processing platform to help decision makers better understand citizen interactions, to estimating electricity consumption through the use of satellite imagery, these projects helped to open up my mind as to how data science can help developing countries to improve service delivery.
- One of the great qualities about Pulse Lab Jakarta’s team is its diversity: the team comprises of passionate data scientists, GIS specialists, humanitarian data advisors and human centered design experts. There’s always an interesting ongoing conversation about big data and thick data at the Lab, and the ways in which they complement each other.
I encountered some challenges when I was running a few data analyses to estimate the electrification rate in Indonesia — ones that data science skills alone could not solve. And that’s because, doing so required additional contextual knowledge of the country and local communities that my analysis of satellite images did not quite reveal. That’s when I began building bridges with the ethnographers in house, who helped me to understand more about the Indonesian archipelago, its regional differences, and how the behaviours of residents in rural and urban areas vary.
I also participated in Pulse Lab Jakarta’s “Research Dive” which is a hackathon-style event that gathered folks from academia as well as the private and the public sectors over four days to analyse given data sets and uncover insights that can be used to address burning policy questions. This year, PLJ collaborated with Indonesia’s Central Bank to understand financial vulnerability at the household level in Indonesia. I participated as an observer in this event, so I got the chance to switch between the different groups of participants who were conducting data analysis to detect customers’ vulnerability based on fintech data and to evaluate the impact that natural hazards have on loans-at-risk. They were composed of different profiles: some coming from the public sector, others from the private sector, some being data-scientists, others university lecturers. Each gave different insights, and this made the analysis more and more interesting and complete.
- Data nowadays is everywhere and we are constantly producing it. This internship introduced me to several open source platforms that have useful data available. One of which is called Humanitarian Data Exchange (HDX), which is an open platform that shares humanitarian data for almost every country, from data on conflict to food security. Thanks to this platform that is managed by the United Nations Office for the Coordination of Humanitarian Affairs, I retrieved shapefiles of all the administrative boundaries of Indonesia, as well as accurate population density, which helped me to spot crowded regions in Indonesia where energy consumption would be presumably higher.
Open Street Map is another platform that provides precise maps with information about roads, trails, amenities and much more. This information can be retrieved in formats that are easy to manipulate for analysis. During my internship, an idea that came up was about how to determine the optimal locations to build schools in Indonesia. For this, we were advised to take into account walkway accessibility to schools and not only look at distance — this is not something I would have automatically thought of with my data science hat on. This was when I was introduced to this amazing platform, and thanks to the data it provides and some python coding applied, we were able to map the pedestrian accessibility to public schools in Jakarta.
- Indonesia is the 4th most populous country, and yet, I knew so little of it and hadn’t had the chance to meet any Indonesians until I travelled there despite living in a cosmopolitan city like Paris. As one Indonesian entrepreneur puts it, Indonesia is probably the biggest invisible thing on earth. The growth of information technology, especially mobile phone usage is exploding in the country and so much data is being created. In terms of data science and data analytics, Indonesia and Southeast Asia in general are thus great places to work! One of Pulse Lab Jakarta’s aims is to help decision makers become more aware of the potential of unconventional data and real-time data.
Unfortunately, around the world many of these data sets are still not being harnessed and analysed as well as they could be. This not only concerns Indonesia, but nearby Southeast Asian countries, and more generally, developing countries. The opportunities are endless: projects involving these new types of data can strongly catalyse South-South cooperation to better consolidate resources, technology, and knowledge between developing countries which can in turn help to realise the Sustainable Development Goals.
To sum things up, I am so grateful for the experience and every person I met in the lab because I learned so much from them in so many aspects. This initiative has already had a positive impact in Indonesia and the region, and I am convinced this will continue. Big Data is something that all governments have to start taking into account to better serve their citizens, and Pulse Lab Jakarta is making headway by building relevant use cases and sharing lessons learned with governments in the region and across the world.
This blog was written by Fabiola Espinoza who interned with Pulse Lab Jakarta during Summer 2019. It has been edited for clarity, format and style consistency.
Pulse Lab Jakarta is grateful for the generous support from the Government of Australia