Through STS (Science and Technology Studies), we we´re to investigate the discourse of the problematization of Smart Cities.
Our investigation had the following research questions:
How can we identify important platforms for discourse on smart cities?
Who are the big (or most influential) actors related to smart city conferences?
How are these actors using smart city discourses to problematize aspects of society?
We used Twitter as a platform to gain knowledge of who we´re talking about Smart Cities. That led to an investigation of different conferences, and the participants at those conferences. This was to see how the participants was problematizing and shaping the discoures on Smart Cities.
The idea was that looking at which topics people tweet about in relation to smart city would provide some sort of measure for an underlying smart city discourse. We (my group) gathered all tweets containing the string smartcity, that were created between August 15 and December 8, 2015, and visualized the network in Gephi. The visualization includes 79.917 tweets containing smartcity.
The below visualization, includes all hashtags being tweeted 100 or more times in relation to smart city during the sample period. This left us with a reasonable number of “topics” (166) related to smart cities to explore.
By manually manipulating the data, we thus made visible patterns appear. The visualization below represents one such map, where hashtags are grouped into clusters of related terms. The visualisation mostly just represent one way for us to make meaning out of the messy data, which in turn allows us to look for interesting “locations” in the communication.
As it turned out, there was a rather significant group of hashtags connected to what appeared to be conferences, congresses, expos and similar events (conferences). In fact, such hashtags were tweeted a total of 8.541 times within the sample of approximately 80.000 total tweets, thereby being represented in more than every 10th smart city related tweet. Being heavily related to these conferences – which are venues for the formation of smart city discourse – can then be thought of as their way of controlling and shaping this discourse.
When we looked into the actors with the heaviest connectivity metrics, four companies stood out as being connected to most smart city conferences: IBM, Accenture, Cisco, and Microsoft. We limited ourselves to include between 5 and 15 of the most relevant texts representing each company. In total, we ended up with 40 texts, comprising well over 300 pages of material. Having thus created a corpus of texts fit for an exploration of the specified discourses, we then used ANTA to analyse the semantic contents.
Colored dots represent individual texts (colored by company). Sizes of expressions represent their frequency relative to length of texts. Weighted edges represent stronger relationships between certain texts and expressions.
Looking at the visualization below we now see a more clear picture after aggregating expressions has thinned out the middle section, making both unique expressions for each company and a general terminology observable.
We find it interesting the way Accenture articulates smart cities. Their discourse uses terms like energy efficiency, fiscal year, technology services, economic development, and high performance delivered, clearly reflecting a very business-oriented view on (smart) cities. As such, Cisco and IBM do not seem to use that unique a vocabulary. Microsoft, standing further from the rest, uses a few, very visible expressions. It is notable how Microsoft articulates their smart city discourse on a very personal and empowering level, as exemplified in expressions like open data, young people, individual citizens, free access, and people-first approach.
I gained knowledge of different theories within STS. This gave me different aspects on how to view a certain topic, controversy or problematization.
Through the project I learned how to use different tools, to either find information or to visualize it, such as Tableau, Gephi, ANTA etc. By knowing different tools to scrape data, I can combine data scrapping tools, and visualize it so it can be simplified to make sense of.