Miguel Ramalho was the big winner of the Arquivo.pt 2020 Award, seeing his work recognized Unarchive – a platform that analyzes news preserved by Arquivo.pt to establish connections between entities, people, and places. The graduate of the Faculty of Engineering of the University of Porto (FEUP) explains the usefulness of this platform, describing the context behind its creation.
The Desarquivo project won 1st Prize in the Arquivo.pt 2020 Award. Can you explain what this solution consists of?
The name "Desarquivo" is partially self-explanatory. It's at least indicative of what it does. It's a project that analyzes archived news text from the last 20 years, identifies references to people, organizations, places, and other entities; creates a network of relationships between these entities based on their mutual presence in a news body; and finally, allows this network—let's call it a graph—to be searched visually and intuitively. It starts with archived news and ends with a graph, a "desarquivo," of the connections between these people, organizations, places, and more.
What are some of the applications of this tool that you see as especially relevant?
The application I initially envisioned is investigative journalism. That's precisely why the focus was on creating a new way to find hidden things. In this case, those things are links. Links that range from discovering that two politicians attended the same event in 2004 to understanding which companies interact most with a given university. Beyond investigative journalism, Desarquivo can be used for exploration, personal learning, and fact-checking. These are applications that allow learning or answering questions, and they have a different nature than what we're used to in conventional search engines. Of course, Desarquivo uses information sources that most search engines don't use, or don't have access to, because they're pages that are no longer present on the web.
Disinformation is seen as one of the defining trends of the 21st century, and its negative impact on our society is often highlighted. Do you believe Desarquivo can help address this phenomenon? How?
The problem of misinformation stems from our widespread human laziness; it's so much easier to scroll and continue reading a feed than to investigate every news story, fact, statistic, or comment we come across. In this sense, I'm not naive enough to think that Desarquivo will change consumption patterns and eliminate idleness in the 21st century. However, I see it having an impact that complements journalism that is truly based on facts and reliable sources of information, and in this sense, it serves as another tool that helps the truth become known, even if it then has to battle for attention with lies. I've also been working on another project aimed at understanding this problem on social media in Portugal, which currently focuses solely on Twitter: Election Watch.
In what context did the basic idea for Desarquivo emerge? And what led you to participate in the Prêmio Arquivo.pt 2020 competition?
The context was precisely a journalistic investigative project by the ICIJ (International Consortium of Investigative Journalists) – the Luanda Leaks – whose documents were later revealed to be the responsibility of Portuguese journalist Rui Pinto. Following this episode, I remember reading an article that explained some of the challenges and approaches used to analyze such a large amount of data and files. From there, and having heard about the Arquivo.pt award, I combined business with pleasure and attempted to tackle a similarly large amount of data, in this case focused on a broader context: Portuguese news content and how it reflects our country. I very much viewed this project as a challenge to do something simultaneously relevant and innovative. Whether it worked or not, I believe time will tell.
Looking to the near future, what do you think the impact of this distinction will be?
I believe I've already experienced a considerable portion of that impact, which included the opportunity to communicate with journalists and share ideas, look at the project from a more technical perspective, and even be able to explain it to audiences and discuss it with people with different backgrounds than myself. I believe there may also soon be some academic interest in improving individual parts of Desarquivo, either from myself or from university students or researchers.
