RÍOS // CHAPTER N.1
RIVERS // AMAZONIA geo-linguistics Visit: https://rivers.ulara.org
Rivers // Amazonia Geo-linguistics is an experimental research tool that maps information related to the Amazon Rainforest on Twitter. The online application explores social media semantics and its potential to identify socioeconomic, political, and environmental issues.The tool has been developed to facilitate the analysis and visualisation of Twitter data, including hashtags and co-occurrences. This data is continuously archived in a database, and algorithms are used to generate statistical data. By analysing the semantics used in tweets, insights can be gained into how Western societies perceive and interact with the complex situation in the Amazon basin.
The results of this analysis are applied to the geo-referenced marks where socio-environmental threads have been identified, performing experimental interventions on the actual topographic data of the Amazonian territory. These interventions result in the sculptures proposed in CHAPTER N.2: SPECULATIVE CARTOGRAPHIES
The starting point of the tool Rivers // Amazonia Geo-linguistics was a carefully selected collection of hashtags related to Amazonia. Based on that initial list of hashtags, we started a database that collects all the hashtags occurrences that appeared in Twitter messages.
For example, #Amazonía is used in the following Tweet, published by AMAZON_NATION on the 23-11-2020:
Territorios Indígenas de la cuenca Amazónica. RÍOS // geo-linguistics // geo-politics // Amazonía work in process #doyourpartfortheamazon #amazonsnations #territorio #amazonia #rios #geopolíticas #mediart #installation #3d #map #mapping #geography
The hashtags present in the Twitter message are included in our database:
Now, these hashtags are part of the database. The application scans Twitter every 30 min. On May 2020 we initiated the database, with an amount of 255 hashtags. Today on January 2021, our database includes 650950 hashtags.
Currently, we perform weekly manual revisions to remove as much noise/irrelevant hashtags from the database. In future development, we will implement machine-learning algorithms that help us maintain an even healthier database.
The database stores the following information:
- Twitter messages
- List of hashtags
- Co-relations between hashtags
- Dates of each Twitter message
- Dates for each hashtag use
- Number of times a hashtag has been used
Users can search for hashtags by introducing a word in the search engine, for example, #deforestation
Users can also use the predefined searches, on top of the page.
The search results are displayed on the following panels:
- Twitter message list: List of messages including the hashtags
- Timeline: Timeline displaying the hashtag used by dates and popularity
- Network diagram (co-occurrece): The network diagram visualizes which hashtags are often used together in tweets by drawing connections between those hashtags and forming clusters.
- Voronoi diagram (vocabulary): The Voronoi (vocabulary) diagram is based on the Voronoi algorithm. This algorithm allows the partition of a plane into regions close to each of a given set of objects. These objects are just points in the plane (called seeds, sites, or generators). For each seed, there is a corresponding region. These regions are called Voronoi cells. Each seed is a hashtag containing information that will define: the size, shape and position of the Voronoi diagram cells.
- The hashtags are paired on a Voronoi Diagram randomly. Each Voronoi graph is unique. For each Voronoi Diagram the position of the cells is translated into X and Y coordinates.
- The number of times a hashtag has been used defines its "popularity". This information determines the size of the cell in the Voronoi diagram. We also translate this information into Z coordinates.
The Voronoi diagram has a series of extra features that we use to create visual interventions onto the geo-referenced marks where socio-environmental threads have been identified, performing experimental interventions on the actual topographic data of the Amazonian territory.
These interventions result in the sculptures proposed in Chapter II: Ríos.
The Voronoi diagram performs the following features:
- Graphic size
- Clip path functions: Each name corresponds to a sub-basin Amazonia.
Next to the clip-path function, there is an arrow icon that allows exporting the current Voronoi diagram in various formats:
- Export Frequencies: Timeline data
- Export co-occurrences: Hashtags co-relations
- Export Geo-Jason: Format designed for representing simple geographical features (spatial coordinates X / Y / Z for each polygon on the diagram), along with non-spatial attributes, like hashtag count.
- Export CSV: Hashtags co-relations and spatial coordinates in CSV format.
- Export STL: the 3D version of the Voronoi diagram. Can be used to visualize the diagram in 3D, where each cell's height corresponds to the number of times a hashtag is used.
- Conception, direction, production, Laura Colmenares Guerra
- Rivers//Amazonia geo-linguistics software, in collaboration with Gijs de Heij
- GIS analyst, Gabriel Codreanu