Social media provide great opportunity to analyse the public opinion related to any product, service or solution. The main goal of the IP4MaaS project, which is to advance the uptake of MaaS schemes, could not be achieved by focusing only on already developed (or soon to be developed) technologies. To truly cater to the needs and wishes of MaaS users, it is crucial to get insights into the public opinion on MaaS. To obtain these insights, IP4MaaS partner Mosaic Factor has performed a so-called sentiment and emotion analysis via social media.
The sentiment and emotion analyses are done by collecting relevant Tweets from Spain (Spanish, Catalan, English) and Italy (Italian, English), and clustering the Tweets under specific ‘most popular’ topic (tickets & prices, schedules, security and COVID19), and analysing them to extract the sentiments and emotions. The overall goal is to see and understand the public comments and feedbacks along with the emotions when they use the public transportation services in some pre-selected areas in Spain (Barcelona and surroundings) and Italy (Padova and surroundings).
The Tweets were collected by searching for specific combinations of key words on public transport, multimodality, public transport companies, among others.
Results
If we look at Spain, the Sentiment Analysis Results are represented below, where the scores vary between -1 (negative) and +1 (positive). In the graph below we can see that overall, more positive feedbacks were given, however the percentage of the negative feedback is non-negligible. Especially when speaking about public transportation in light of the COVID-19 crisis, the negativity is very close to the positivity levels. This was to be expected, considering most feedbacks were given during the adaptation processes and it is possible that passengers complained on social media about safety regulations, restrictions, etc. Another significant result is the ‘ticket’ topic: this has much more positive feedbacks than the other topics. This can be explained by the ongoing advancements in both cities concerning new ticketing and payment services, innovations which are in general beneficial for passengers.
The emotion analyses results shown that the “Schedule” topic is the one that the users connect most to the “anger” emotion, while at the same time they have less trust into the topic. This leads to consider that the current schedules or reliability of the schedules are problematic in Spain. On the contrary, “trust” is the predominant sentiment, which can represent the overall trust to the public transportation system.
With only two topics in Italy, again the overall trust into the public transportation system is significant, whereas “anticipation” is the most frequent sentiment. The latter can be explained by the knowledge of the service by the users, regarding the topics. The “sadness” emotion in the “Schedule” topic again shows that there might also be problems with scheduling and reliability in Italy.
The outcomes from the sentiment and emotion analyses complement the conversational survey results (more about that here!), providing feedback to IP4 solution providers, as well as informing the actual IP4MaaS experimentation.