BLOG – Analysis

#SmartCity : What Twitter says

In recent years, the topic of Smart Cities has been increasingly discussed. In order to know more about what the public opinion has to say about this concept, our post-doc researcher, Jessica Clement, took a closer look at the use of #SmartCity on Twitter : a well-known and influencial social network, but also an efficient monitoring tool.

So, this time we invite you to take a completely different path: Take a dive with us into the world of Twitter and discover our researcher’s observations.
#Smartcity Image large 

Twitter and #SmartCity 

Twitter, an American social network launched in 2006, allows its users, through short and purposeful messages – commonly called Tweets – to share opinions on many topics. In 2021 the company claims it counts more than 200 million daily active users, mainly among the 18 to 49 age group (79%).

For years, hashtags – recognizable by the “#” sign placed in front of a word or a series of words – have been used on social networks to link information to a subject. They are widely used on Twitter but also on other networks such as Instagram and Facebook.

This is why Jessica Clement naturally suggested to analyze Tweets mentioning #SmartCity. However, why should we be interested in Twitter more than any other platform?

First, from a practical perspective, Twitter supports research by allowing researchers to use its data. This simplifies the information gathering process and has also led to cutting-edge studies on Smart Cities. As an example, an Australian scientific paper published in 2021 aimed to analyze the perceptions and uses of Smart City concepts and technologies on Twitter (Yigitcanlar and al., 2021).

But this is not the only reason. The platform’s users tend to leverage the network to discuss topical issues and to express their feelings and opinions through their Tweets. The study of a large sample of these short messages thus becomes an efficient way to understand the main topics and sentiments related to a theme, even when it is evoked in different contexts.

Key figures

As the term Smart City is frequently used around the globe, to develop her study, Jessica Clement was interested in Tweets published in both French and English between June 2020 and May 2021. In this way she was able to see if similarities could be observed based on the language used*.

83 100

This is the number of Tweets mentioning #SmartCity that could be collected over the study period. This represents an average of 228 Tweets a day or 9 Tweets an hour. Among these 83,100 Tweets, 50,187 (60%) were in English and 6,554 (8%) in French. By way of comparison, every second, about 9,120 Tweets are shared (more than 30 million an hour) and 125 million hashtags are used every day. By 2020, #Covid19 was the most popular hashtag in the world with nearly 400 million mentions on the platform or nearly 760 Tweets a minute.
Note that in order to get geolocation data we have to know Twitter users’ location, which can be added to their profile or any given Tweet. As this is not regularly practiced, the study of two languages proved to be more meaningful than the analysis of Tweets based on countries or regions of origin. Fortunately, the English term Smart City is widely used in all languages.

#SmartCity : Tweets mainly focusing on technology?

Thanks to the tool made available by the American social network, it is possible to identify the terms – and therefore the subjects – most frequently associated with a hashtag.

Regarding #SmartCity, we notice that depending on the language of the Tweets, some parallels can be observed. Indeed, the Internet of Things (IoT) seems to occupy a major place in both wordclouds (Figure 1). As a reminder, the IoT refers to all the connected objects using sensors, software or other technologies. Other terms such as data, big data, technology, digital (or numérique in French), AI (Artificial Intelligence), 5G, etc. enrich the clouds containing the terms that are the most commonly attached to the subject matter. It leads us to the conclusion that, globally, Smart Cities are mainly related to the technological aspect of the concept.

#smartcity nuage mots 
Figure 1 : Wordclouds of terms that are the most commonly associated to #SmartCity in French and English Tweets.
In both languages the idea of future and evolution is also present with words like demain (which means “tomorrow” in French) and future, but also innovation, solution and project (projet in French).

Nevertheless, it should be pointed out that despite a common association of Smart Cities to technologies – which seems quite logical as they facilitate territorial initiatives – the French-speaking vision appears to take more account of sustainability aspects. For example, we can observe the appearance of the terms durable (sustainable) and environnement (environment) alongside territoire (territory) and service.


#SmartCity : Positive Tweets overall

A sentiment analysis, using the Tweets provided by Twitter, allowed our researcher to observe that, on average, the Tweets presenting the #SmartCity were positively written. As shown in Figure 2, 63% of the Tweets in French and 68% of the Tweets in English were deemed positive compared to 7% of negative messages in both languages.  
#smartcity Analyse sentiments Global EN 
Figure 2 : Tweet sentiments (positive, neutral or negative) mentioning #SmartCity according to the language used. 

In order to bring this analysis into perspective, the same exercise was conducted for Tweets presenting #technology (#technologie for French Tweets) between June 2020 and May 2021. Indeed, as technology (as such or through associated concepts) is often related to #SmartCity, it seemed relevant to understand how it was globally perceived on the social network.

In French, 57% of the Tweets mentioning #technologie were positive, while nearly 10% were more negative. In English (using #technology) the rates were respectively 60 and 11%. In both languages, Tweets explicitly presenting #SmartCity are therefore written in a slightly more positive and less negative way.


#SmartCity : Related themes

Thanks to her analysis, Jessica Clement was able to identify for each language the 3 most frequently referred to topics when it comes to Smart Cities. 
iconeInfo How are these main subjects determined? This “categorization” is done through a topic modeling analysis on each set of Tweets. In other words, this analysis finds “hidden” topics in a large group of documents such as Tweets. Topics are characterized by a set of words that are most relevant to each defined topic.

Tweets in French : Focus on Smart City challenges

The observation of the modeling results allowed Jessica Clement to identify 3 main topics related to #SmartCity among the French Tweets :
  1. The first topic to emerge in the messages posted on the network is characterized by keywords such as start-up, demain, capteur (“sensor”), challenge, solution, smart-building and projet. This topic therefore encompasses a process of reflection and of territorial innovation, elements that are at the core of sustainable and smart territories. 
  2. The second most frequent theme rather targeted challenges related to mobility, whether environmental or technological. Among the keywords associated with this topic : mobilité (“mobility”), transport, territoires (“territories”), données (“data”) and environnement (“environment”).
  3. Finally, the third subject identified is more focused on environmental and energy challenges. Indeed the following words stand out in this category : énergie (“energy”), environnement, durable (“sustainable”), solution, capteurs, IdO (IoT) et numérique (“digital”).
The sentiment analysis shows that these three topics are usually positively represented in the Tweets studied. 
#smartcity Analyse sentiments Sujets FR graphEN 
Figure 3 : Sentiments associated with the three topics derived from the French Tweets.

Tweets in English : A more technological dimension

The three main subjects were also identified for English Tweets :
  1. First, we also find the aspect of reflection and innovation for the future, however there is a more marked technological dimension here than with the French Tweets. Therefore, we find words like innovation, data, solution, technology, new and future. It is also interesting to note that in English, the term Covid appears regularly, as it is among the 30 most frequently used termsOur researcher, Jessica Clement, comments, “As the Tweets were gathered in the context of the pandemic, the Smart City seems to be envisioned as a way to cope with the challenges that accompany Covid-19 and to initiate a renewal in our territories. Indeed, this aligns with recent research conducted by the Smart City Institute, suggesting that the transition to sustainable and smart cities can be coupled with the recovery from the health crisis (Clement & Crutzen, 2021).
  2. Next, it is the theme of autonomous/driverless vehicles that most often comes up. This theme has the words autonomous, self-driving, and driverless, as well as the technologies underlying this new form of mobility like big-data, machine-learning, robotics, and IoT.
  3. Finally, the third most common topic was the Internet of Things (IoT). It is naturally attached to keywords like IoT, technology, IIoT (Industrial Internet of Things), IoTPL (IoT Premier League), IoTcommunity, and SmartThings.
The sentiment analysis reveals that the topic of driverless vehicles seems to be the most controversial. Indeed, this category presents both the lowest percentage of Tweets with a positive sentiment (45%) and the highest percentage with negative feelings (8%). Nevertheless, we should also notice the high proportion of Tweets that presented the information in a neutral way. 
#smartcity Analyse sentiments Sujets EN Figure 4 : Sentiment associated with the three topics derived from the English Tweets.

What can we learn from our immersion into the realm of Twitter?  

After scouring Twitter and several tens of thousands of Tweets, we can say that, on the whole, the Smart City concept is mostly discussed in a positive manner. This is seen regardless of the language used by tweeters.

Another similarity between the two languages is that technology seems to play an important role in the messages conveyed about the Smart City in the Twitterverse. Hence, technology still seems to underlie all Smart Cities discussions.

However, the way in which technology is approached appears to differ slightly between the French and English perspectives. In the French Tweets, the Smart Cities challenges are more often evoked and the possible technological solutions are associated with them. Technology is therefore not very prominent in the messages despite its undeniable presence. On the other hand, Tweets in English seem to deal with the Smart City in more tangible and technical terms, seen through their frequent use of terms like IoT, technology, startup, big data and innovation. In this way, the discussions seem to revolve more around the technological solutions identified.

Considering that some studies predict the growth of investments in technological solutions to territorial challenges, it is likely that discussions on the subject are destined to increase. This suggests to our researcher, in conclusion, that this kind of study can undoubtedly help to better understand what can be expected in the future.

Sources and references – For more information :
Clement, J., & Crutzen, N. (2021). COVID-19: An Accelerator for Sustainable and Smart Urban Transitions? Georgetown Journal of International Affairs.
Yigitcanlar, T., Kankanamge, N., & Vella, K. (2021). How Are Smart City Concepts and Technologies Perceived and Utilized? A Systematic Geo-Twitter Analysis of Smart Cities in Australia. Journal of Urban Technology, 28(1–2), 135–154. 

Photo credit : Robin Worrall - Unsplash

Share this news