The project
OBSERVATÓR!O2016 is an online platform for collecting, visualizing and analysing social data, news feeds and statistics concerning the Olympics, which will be held in Rio de Janeiro in August 2016.
The Olympic Games is a worldwide event that gathers attention from both media and online audience. As a consequence, this sporting event becomes a social phenomena that produces large data streams from different sources and stakeholders. The most accessible and richiest source of data explored in this project is Twitter messages (text and media), collected via Twitter's public API.
Experimental concept
The project is part of a Ph.D thesis interested in investigating a methodological framework for visualizing cities phenomenas as complex informational landscapes where multiple layers of "digital traces" coexist. At the same time, OBSERVATÓR!O2016 is also part of VISGRAF a research Laboratory that, among another activities, is concerned about exploring different informational technologies for collecting, parsing and representing data.
Saying that, we conceive the project as an experimental work in which the process of creation, development and learning is so important as the final result. That's why OBSERVATÓR!O2016 is launched a couple of months before the Opening Ceremony. Our goal is to iterate the project several times so we can learn from the process and refine OBSERVATÓR!O2016's concept and techniques.
UPDATE A more detailed report of the project can be read in OBSERVATÓR!O2016 documentation page.
Web portal beta version
The beta version of the Web Portal was launched on March 2016 and focused on bringing together tweets (text and media) and news exclusively from the Olympics “official" source, that is the twitter account and website of the Rio 2016 Olympics. The temporal apprehension of the experiments designed at this stage was essentially static and inert, snapshots that captureded a especfic moment of the Pre Olympics.
Web portal version 1.0
As the project unfolded and gained life, we started to combine the "official" data of the Olympics Games with data streams shared by other group of actors with different interests and perspectives, such as ordinary user accounts, critical spokespersons and both mainstream and alternative media. These multiple perceptions combined provide a broad observatory on how the Olympics Games are experienced and shared, from its preparation to its conclusion.
The broader and more heterogeneous approach at this project stage was followed by experiments that reach a temporal dimension. The visualizations here trie to show a dynamic and dialetic conceptualisation of the opinion shared about Rio 2016 Olympics Games.
Web portal version 2.0
In order to enhance the portal for the Games, we considered interesting aspects of the Olympics that would be in evidence during the competitions. This led our project to focus on sporting disciplines, nations and individual athletes. We explored more sophisticated visualizations capable of communicating layered meanings and improved some features from the older versions.
Finally, at this stage the new experiences incorporate an almost real-time fruition of the data.
Web portal version 3.0
During the Games, our work focuses on three main directions: following the topics that were generating more buzz to feed our blog with stories; identifing athletes with better performance to create a new visualization; implementing temporal handlers in visualisations that have not yet presented this feature.
Future work
After the Olympics, the projetct will continue to expand. OBSERVATÓR!O2016 portal will be a digital memory of commentary and opinion on Rio-2016, shared on Twitter. Since the data collected are stored in their entirety, future data analysis and visualization are possible. In this sense, a partnership between VISGRAF and the Computing Institute of UFF (Federal Fluminense University) is ongoing and involves deep learning concepts and techniques from the collected data about the Olympic Games.
UPDATE We have just released an API Rest for reading the results of our image classification using a deep learning model. For a full API access please contact us on observatorioolimpico@gmail.com