Sentiment Analysis of Social Media Use in Public Transportation in

Sweden

Azadeh Sarkheyli

and Elnaz Sarkheyli

School of Information Technology, Halmstad University, Halmstad, Sweden

Institute for Urban Studies, Department of Urban Studies, Malmö University, Malmö, Sweden

Keywords: Social Media, Public Transportation, Sentiment Analysis, Sweden.

Abstract: The increasing impact of social media on public transportation is transforming communication strategies and

user engagement. These platforms offer real-time service updates while allowing users to voice their concerns

and suggestions, fostering trust and enhancing customer satisfaction. This research investigates public

perceptions of the communication methods used by public transportation services in Sweden, highlighting

user preferences for different social media platforms and content types. The study employs a four-step

methodology. First, a literature review examines the advantages and challenges of integrating social media

into public transportation systems. Next, a survey assesses Swedish users’ interactions with social media in

this context. The third phase involves sentiment analysis and text mining of the survey responses to evaluate

public opinion. Finally, the research proposes potential steps for collecting and analyzing social media data.

The findings contribute to a better understanding of effective communication strategies, ultimately improving

the responsiveness of public transportation systems.

1 INTRODUCTION

Sentiment analysis utilizes natural language

processing and machine learning techniques to

identify and extract subjective information from text,

allowing for the classification of sentiments as

positive, negative, or neutral. This method has gained

significant popularity due to its capacity to analyze

vast quantities of unstructured data derived from

sources such as social media (Yue et al., 2019).

Social media analytics encompasses the collection

and examination of data from social media platforms

to derive insights into user behavior and emerging

trends. This process includes monitoring interactions

and evaluating content performance, which can

ultimately influence and shape public perception

(Rodríguez-Ibánez et al., 2023; Chowdhury, 2024).

According to Rodríguez-Ibánez et al. (2023),

sentiment analysis has become a prominent research

area since 2008, as evidenced by the increasing

number of published studies. The research focus has

varied widely, encompassing emergencies, business

intelligence, marketing, and the prediction of

https://orcid.org/0000-0002-5390-7509

https://orcid.org/0000-0003-2618-8796

electoral outcomes.

Consequently, there are numerous examples of

sentiment analysis within social media, including the

work by Brahami et al. (2025), who developed a

sentiment analysis model using Knowledge Graph

Convolutional Networks (KGCN) to analyze over

410,000 tweets related to the Russia–Ukraine

conflict. This model achieved an accuracy of around

89%, demonstrating that the incorporation of

knowledge graphs significantly enhances sentiment

classification. This method also offers promising

potential for real-time monitoring of public opinion,

with substantial implications for policy. Another

example is the study by Salisu et al. (2024), which

examined sentiments expressed in comments on the

official Instagram account of Universitas

Widyagama. Their findings indicate effective

promotion of a positive image through social media

and highlight the potential of sentiment analysis for

strategic communication in higher education.

In addition, Abeysekera (2024) studies customer

attitudes toward a pop-up tax clinic in Australia that

helps disadvantaged taxpayers. Analysis of feedback

522

Sarkheyli, A. and Sarkheyli, E.

Sentiment Analysis of Social Media Use in Public Transportation in Sweden.

DOI: 10.5220/0013821100004000

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 17th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2025) - Volume 2: KEOD and KMIS, pages

522-528

ISBN: 978-989-758-769-6; ISSN: 2184-3228

from 47 clients showed that free services and good

customer care improved sentiment, while factors like

tax year and location had little impact. The findings

indicate the potential of such clinics to reduce

inequality and offer a new method for evaluating

social enterprises through sentiment analysis.

A pertinent example relevant to this research is the

study conducted by Torres and de Picado-Santos

(2025), which examines 81 open-access research

papers published between 2014 and 2024 that focus

on sentiment analysis and topic modeling within the

field of transportation. It underscores the success of

these techniques, particularly when applied to social

media data, in gaining insights into user attitudes and

fostering sustainable mobility. Significant obstacles

include linguistic variety and the diversity of data

types, prompting the authors to suggest a framework

for selecting studies and performing bibliometric

analysis. The paper points out the opportunities

presented by integrating these methodologies to guide

the development of more intelligent and inclusive

transportation policies.

Therefore, the role of social media in public

transportation is increasingly recognized as vital for

improving communication and user engagement.

These platforms facilitate real-time updates on

service statuses and allow users to voice concerns and

suggestions quickly (Aman & Smith-Colin, 2021; Liu

& Ban, 2017).

Social media creates a dynamic environment for

two-way communication between transit authorities

and the public, fostering trust and enhancing

customer satisfaction by addressing user feedback

directly (Liu & Ben, 2017; Zeng & Gerritsen, 2014;

Kwok & Yu, 2013).

Additionally, engaging users on these platforms

encourages community involvement and a sense of

ownership over services (Han et al., 2020). This study

explores public sentiment toward communication

strategies used by public transportation services in

Sweden, analyzing user preferences for social media

platforms and content types that resonate most.

Understanding these aspects is crucial for enhancing

communication strategies and creating a more

responsive public transportation system (Sarkheyli &

Sarkheyli, 2024; Han et al., 2020; Bergman, 2012).

This study is structured into four phases. In the

first phase, we conducted a literature review to

examine insights into social media usage in research,

particularly examining the advantages and

disadvantages of employing social media within the

scope of public transportation. The second phase

involved a survey to understand how individuals in

Sweden utilize social media for public transportation,

serving as a medium-scale study. In the third phase,

we performed sentiment analysis and text mining on

the survey comments to evaluate public sentiment

regarding the use of social media for public

transportation in Sweden. The fourth phase, which is

related to social media data collection and analysis, is

mentioned for the next step of this research. It is

informed by the findings from the preceding phases

(refer to Figure 1).

2 SOCIAL MEDIA AND PUBLIC

TRANSPORTATION IN

SWEDEN

Social media has become an integral tool for public

transportation agencies in Sweden and Europe to

engage with passengers and improve services. Transit

operators primarily use platforms like Facebook and

X for real-time updates, service information, and

addressing customer concerns (Sarkheyli &

Sarkheyli, 2024; Georgiadis et al., 2020). A survey of

European agencies found Twitter (or X as its new

name) most effective for short communications and

service updates. At the same time, Facebook was

valid for announcements and community building

(Georgiadis et al., 2020).

Transit agencies in the United States and Canada

have adopted social media for timely updates, public

service, and citizen engagement (Bregman, 2012).

During unplanned rail disruptions, social media,

particularly Twitter, proves valuable for providing

concise, real-time information to passengers.

However, challenges remain, including staff

resourcing and managing passenger expectations

(Nikolaidou & Papaioannou, 2018). Social media

integration in public transportation offers new

opportunities for improving urban mobility systems

(Sarkheyli & Sarkheyli, 2024).

According to the SWOT analysis of using social

media in public transportation, social media offers

public transportation agencies valuable advantages

such as data access, direct passenger communication,

and cost-effective service promotion, enhancing

customer satisfaction and loyalty.

However, it requires resources and presents

challenges like data privacy issues, negative

feedback, and limited accessibility. Opportunities for

improvement include leveraging passenger feedback

and building community partnerships, while threats

consist of cybersecurity risks, evolving regulations,

and competition from ride-sharing services. Agencies

need to strategically manage these factors to

Sentiment Analysis of Social Media Use in Public Transportation in Sweden

523

maximize benefits (Sarkheyli & Sarkheyli, 2024;

Kaplan & Haenlein, 2010; Mangold & Faulds, 2009;

Munar & Jacobsen, 2014).

3 RESEARCH METHODOLOGY

The research has been organized into four phases. In

Phase I, a literature review was conducted to

understand the main concepts of the study. Phase II

involved distributing an online questionnaire to

gather insights on how individuals in Sweden utilize

social media platforms for public transportation. The

survey targeted a diverse group of respondents,

including students, teachers, researchers, engineers,

doctors, and other professionals living and working in

Sweden, to capture a different range of perspectives.

In Phase III, the data collected from the survey

underwent a thorough analysis using sentiment

analysis and text-mining techniques to uncover

patterns and sentiments expressed in the responses.

Each response was categorized into three sentiment

classes: positive, negative, or neutral. Finally, Phase

IV analyzes social media platforms identified in the

earlier phases. This paper presents results from

Phases I, II, and III, specifically the sentiment

analysis, as illustrated in Figure 1.

Figure 1: The research process.

The sentiment analysis utilized two tools:

VADER (Valence Aware Dictionary and sEntiment

Reasoner), which employs a lexicon-based approach,

and XLSTAT, which incorporates statistical and

machine learning techniques (Yordanova et al.,

2021). VADER is ideal for sentiment analysis as it

effectively handles social media nuances like slang

and emoticons, which traditional methods often miss.

It uses a scoring system to assess positive, negative,

and neutral terms, providing accurate and fast

analyses (Dixit et al., 2025).

The research combines XLSTAT and VADER for

sentiment analysis, enhancing accuracy with user

feedback on public transportation. Key themes were

identified and visualized in charts for clarity and

comparison.

4 RESULTS AND DISCUSSION

An online survey gathered 106 responses, with

approximately 49% females and 46% males, averaging

37 years of age. The educational attainment showed

that 36% held Doctorates or University/College

degrees, 21% had Master's degrees, 4% were High

School graduates, 2% completed Vocational training,

and 1% fell into the other category. In terms of

geographical distribution in Sweden, Skåne accounted

for 71% of the responses, followed by Halland at 20%,

while Stockholm, Västra Götaland, Kronobergs, and

Östergötland represented smaller shares of 4%, 3%,

1%, and 1%, respectively (see Figure 2).

Figure 2: The city/region of the respondents.

Figure 3 illustrates the use of various social media

platforms for communication purposes. Instagram is

the most frequently used platform, with a value

slightly above 60, followed closely by Facebook, also

above 60. LinkedIn is the third most utilized platform,

with usage close to 50. Other types of social media

have moderate use, reaching around 30, while X

(formerly Twitter) is used less, with a value of 20.

TikTok has the lowest usage for communication

purposes, with a value below 10.

KMIS 2025 - 17th International Conference on Knowledge Management and Information Systems

524

The chart indicates that visual and text-based

platforms like Instagram and Facebook are the most

preferred for communication, while TikTok, known

for short-form videos, is the least used in this context.

Figure 3: Use of social media for communication purposes.

4.1 Social Media Usage for Public

Transport

Figure 4 shows the distribution of social media usage

for public transport-related purposes across different

platforms. Facebook dominates with 59%, indicating

that it is the most used platform for sharing and

accessing public transport information, updates, or

discussions. Instagram follows with 17%, suggesting

moderate engagement, possibly for visual content

related to public transport. Both X (formerly Twitter)

and LinkedIn have equal shares of 12%, indicating a

lower but still notable usage, likely for real-time

updates (X) and professional discussions or policy-

related content (LinkedIn).

The chart highlights that Facebook is the primary

platform for public transport engagement, while other

social media channels have significantly smaller

shares. Regarding social media engagement, the

survey found that 54% of respondents expressed

dissatisfaction with public transportation. In contrast,

14% of posts or comments conveyed appreciation,

another 14% were focused on providing information,

7% were recommendations, and the remaining 11%

encompassed a range of other topics (see Figure 5).

As highlighted by the survey results (refer to

Figure 6), nearly half of the respondents, precisely

49%, reported not participating in posting or

commenting on social media. Conversely, a smaller

percentage, precisely 20% of the respondents,

indicated they were actively engaged on various

social media platforms.

Figure 4: Usage of social media platforms for public

transportation.

Figure 5: Category of social media posts and comments

about public transportation by people.

Figure 6: Engaged in posting and/or commenting on social

media platforms about public transportation.

4.2 Sentiment Analysis

The sentiment analysis performed with XLSTAT

uncovered a spectrum of feedback from respondents

Sentiment Analysis of Social Media Use in Public Transportation in Sweden

525

about their experiences. Significantly, 60% of the

analyzed comments reflected positive sentiments,

with many users expressing appreciation for the

timely information updates and real-time alerts.

Conversely, 30% of respondents reported

dissatisfaction, citing concerns over outdated

information and a perceived lack of responsiveness.

Finally, 10% of the responses were categorized as

neutral, providing general feedback without strong

emotional overtones.

The comment word cloud is illustrated in Figure

7, while Figure 8 displays the document sentiment

scores. Additionally, Figure 9 presents the

distribution of document scores, and Figure 10 shows

term frequencies, indicating that a term's frequency

increases with the number of times it appears in the

documents.

Figure 11 shows the sentiment-based word cloud,

where the larger a term is in the word cloud, the more

frequent the term is in the document. Each color

represents an emotion. In addition, Figure 12 shows

the term scores, showing that the more positive a term

is, the higher its sentiment score will be.

Figure 7: Word cloud of the comments.

Figure 8: Document sentiment scores.

Figure 9: Document scores distribution.

Figure 10: Term frequencies.

Figure 11: Sentiment-based word cloud.

-2

-1

X10

X11

X12

X13

X14

X15

X16

X17

X18

X19

X20

X21

X22

X23

X24

X25

X26

X27

X28

X29

X30

X31

X32

X33

X34

X35

X36

X37

Document

Sentiment score

Score

-2

-1

Document sentiment scores

-2 0 2

Document score

Frequency

Document scores distribution

make

sk.netrafiken

good

work

delay

info

time

updat

platform

app

inform

media

public

social

transport

051015

Frequency

Te r m

Frequency

5.0

7.5

10.0

12.5

15.0

Term frequencies

positive

delay

good

work

disrupt

ruin

clear

free

great

addict

critic

difficult

disturb

fail

lack

limit

mess

miser

miss

shame

threat

wrong

cheaper

clean

cool

correct

easier

famous

luck

nice

prefer

quicker

safe

super

support

wonder

negative

positive

KMIS 2025 - 17th International Conference on Knowledge Management and Information Systems

526

Figure 12: Term scores.

Figures 13 and 14 display the results of the

VADER sentiment analysis conducted on the textual

comments. The pie chart (Figure 13) illustrates the

distribution of comments' sentiment categories.

Positive sentiment comprises 62%, showing a clear

majority, while 24% is classified as neutral,

indicating a lack of strong feelings. Only 14% of the

data reflects negative sentiment, representing a

smaller portion. Overall, the chart highlights the

dominance of positive sentiments, with neutral and

negative sentiments being less prevalent. In addition,

Figure 14 shows that most sentiment scores are

positive, with values predominantly clustering

between 0.4 and 0.8.

Figure 13: VADER sentiment analysis result.

However, a few instances, specifically at indices

around 9, 25, 29, and 31, exhibit negative sentiment

scores, indicating some negative sentiment. The data

reflects fluctuations in sentiment, with occasional

dips into negative values, yet there is a clear overall

trend toward positive sentiment.

Figure 14: VADER sentiment scores.

4.3 Common Themes in Feedback

In analyzing the feedback, several key themes surfaced

prominently. Many respondents underscored the

importance of information accuracy, expressing a

strong desire for more reliable and timely updates

concerning transport services. Additionally, there was a

clear demand for enhanced engagement and interaction,

as users wished for more direct communication from

transport authorities on social media platforms.

Furthermore, the need for real-time alerts was

emphasized, with numerous users noting that instant

notifications regarding delays and disruptions would

greatly improve their overall experience.

The satisfaction levels among respondents were

notably varied. Approximately 40% reported feeling

neutral or having no strong opinion. Meanwhile, 35%

expressed dissatisfaction due to outdated or

inadequate information. On a more positive note,

25% of respondents indicated that they were satisfied

with the current state of social media communication.

5 CONCLUSIONS

Recent advancements in social media have made it

essential for companies to connect with a broader

audience and improve relationships.

This research examines how the public

transportation sector in Sweden uses social media to

enhance communication and services, structured into

four phases (see Figure 1). This paper presents survey

results and conducts sentiment analysis on comments

relevant to Phases I, II, and III.

The first phase involved a literature review on the

use of social media and its regulations in Sweden,

along with relevant studies on its application in public

transport. The second and third phases included a

survey of 106 Swedish residents to evaluate their use

of social media for public transportation.

The results indicated that apps like SJ and

Skånetrafiken are the most popular sources for

addict

critic

delay

difficult

disrupt

disturb

fail

lack

limit

mess

miser

miss

ruin

shame

threat

wrong

cheaper

clean

clear

cool

correct

easier

famous

free

good

great

luck

nice

prefer

quicker

safe

super

support

wonder

work

Sentiment score

Ter m

negative

positive

Te rm s co res

Sentiment Analysis of Social Media Use in Public Transportation in Sweden

527

transportation updates and information. A strong

preference for receiving updates through social media

was noted, particularly concerning service changes

and travel disruptions.

The study emphasizes that while social media

plays a crucial role in communicating about public

transport, there is considerable room for improvement

in terms of information accuracy, user engagement,

and real-time updates. Public transport authorities

should enhance their digital strategies by improving

the responsiveness of their social media channels and

ensuring timely, accurate information. Proactive

engagement with users can also enrich the passenger

experience, fostering a more informed ridership and

increasing overall satisfaction.

ACKNOWLEDGEMENTS

This paper is part of a research project titled

"Exploring Citizen Satisfaction with Public

Transportation through Social Media and Open Text

Surveys." The project is conducted by Malmö

University in collaboration with Halmstad University

and has been supported and funded by K2 (The

Swedish Knowledge Centre for Public Transport).

REFERENCES

Abeysekera, I. (2024). Sentiment analysis of a nomadic tax

social enterprise. Journal of Open Innovation:

Technology, Market, and Complexity, 10(1), 100179.

Aman, J. J., & Smith-Colin, J. (2021). Leveraging social

media to understand public perceptions toward

micromobility policies: The dallas scooter ban case.

Findings.

Bregman, S. (2012). Uses of social media in public

transportation (No. 99). Transportation Research Board.

Chowdhury, R. H. (2024). Sentiment analysis and social

media analytics in brand management: Techniques,

trends, and implications. World Journal of Advanced

Research and Reviews, 23(2), 287-296.

Dixit, M., Bali, M. S., Kour, K., Ioannou, I., Ghantasala, G.

P., & Vassiliou, V. (2025). Video sentiment analysis on

social media using an advanced VADER technique.

International Journal of Data Science and Analytics, 1-33.

Georgiadis, G., Nikolaidou, A., Politis, I., & Papaioannou,

P. (2020). How public transport could benefit from

social media? evidence from european agencies. In

Conference on Sustainable Urban Mobility (pp. 645-

653). Cham: Springer International Publishing.

Han, X., Wang, J., Zhang, M., & Wang, X. (2020). Using

social media to mine and analyze public opinion related

to COVID-19 in China. International journal of

environmental research and public health, 17(8), 2788.

Kaplan, A. M., & Haenlein, M. (2010). Users of the world,

unite! The challenges and opportunities of Social

Media. Business Horizons, 53(1), 59-68.

Kwok, L., & Yu, B. (2013). Spreading social media

messages on Facebook: An analysis of restaurant

business-to-consumer communications. Cornell

Hospitality Quarterly, 54(1), 84-94.

Liu, J. H. I., & Ban, X. J. (2017). Measuring the impacts of

social media on advancing public transit (No. NITC-

RR-654). National Institute for Transportation and

Communities.

Mangold, W. G., & Faulds, D. J. (2009). Social media: The

new hybrid element of the promotion mix. Business

Horizons, 52(4), 357-365.

Menaouer, B., Fairouz, S., Meriem, M. B., Mohammed, S.,

& Nada, M. (2025). A sentiment analysis of the

Ukraine-Russia War tweets using knowledge graph

convolutional networks. International Journal of

Information Technology, 1-18.

Munar, A. M., & Jacobsen, J. K. S. (2014). Motivations for

sharing tourism experiences through social media.

Tourism Management, 43, 46-54.

Nikolaidou, A., & Papaioannou, P. (2018). Utilizing social

media in transport planning and public transit quality:

Survey of literature. Journal of Transportation

Engineering, Part A: Systems, 144(4), 04018007.

Rodríguez-Ibánez, M., Casánez-Ventura, A., Castejón-

Mateos, F., & Cuenca-Jiménez, P. M. (2023). A review

on sentiment analysis from social media platforms.

Expert Systems with Applications, 223, 119862.

Salisu, I. A., Ramadhan, I. R., Matdoan, S., Arifin, Z., &

Praseptiawan, M. (2024). Sentiment Analysis of

Comments on Higher Education Social Media Using

Naïve Bayes Algorithm. Journal of Information

Technology application in Education, Economy, Health

and Agriculture, 1(3), 57-62.

Sarkheyli, A., & Sarkheyli, E. (2024). Social Media

Integration in Public Transportation: A Case Study of

Sweden. In 2024 IEEE International Conferences on

Internet of Things (iThings) and IEEE Green

Computing & Communications (GreenCom) and IEEE

Cyber, Physical & Social Computing (CPSCom) and

IEEE Smart Data (SmartData) and IEEE Congress on

Cybermatics (pp. 595-600). IEEE.

Torres, E. C. M., & de Picado-Santos, L. G. (2025).

Sentiment Analysis and Topic Modeling in

Transportation: A Literature Review. Applied Sciences,

15(12), 6576.

Yordanova, L., Kiryakova, G., Veleva, P., Angelova, N., &

Yordanova, A. (2021). Criteria for selection of

statistical data processing software. In IOP Conference

Series: Materials Science and Engineering (Vol. 1031,

No. 1, p. 012067). IOP Publishing.

Yue, L., Chen, W., Li, X., Zuo, W., & Yin, M. (2019). A

survey of sentiment analysis in social media.

Knowledge and information systems, 60(2), 617-663.

Zeng, B., & Gerritsen, R. (2014). What do we know about

social media in tourism? A review. Tourism

Management Perspectives, 10, 27-36.

KMIS 2025 - 17th International Conference on Knowledge Management and Information Systems

528